id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.7k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
|
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1309.4411 | Ginestra Bianconi | Arda Halu, Satyam Mukherjee and Ginestra Bianconi | Emergence of overlap in ensembles of spatial multiplexes and statistical
mechanics of spatial interacting networks ensembles | (12 pages, 4 figures) for downloading data see URL
http://sites.google.com/site/satyammukherjee/pubs | Phys. Rev. E 89, 012806 (2014) | 10.1103/PhysRevE.89.012806 | null | physics.soc-ph cond-mat.dis-nn cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spatial networks range from the brain networks, to transportation networks
and infrastructures. Recently interacting and multiplex networks are attracting
great attention because their dynamics and robustness cannot be understood
without treating at the same time several networks. Here we present maximal
entropy ensembles of spatial multiplex and spatial interacting networks that
can be used in order to model spatial multilayer network structures and to
build null models of real datasets. We show that spatial multiplex naturally
develop a significant overlap of the links, a noticeable property of many
multiplexes that can affect significantly the dynamics taking place on them.
Additionally, we characterize ensembles of spatial interacting networks and we
analyse the structure of interacting airport and railway networks in India,
showing the effect of space in determining the link probability.
| [
{
"version": "v1",
"created": "Tue, 17 Sep 2013 18:05:29 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Dec 2013 17:59:58 GMT"
},
{
"version": "v3",
"created": "Wed, 29 Apr 2015 19:49:40 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Halu",
"Arda",
""
],
[
"Mukherjee",
"Satyam",
""
],
[
"Bianconi",
"Ginestra",
""
]
] | TITLE: Emergence of overlap in ensembles of spatial multiplexes and statistical
mechanics of spatial interacting networks ensembles
ABSTRACT: Spatial networks range from the brain networks, to transportation networks
and infrastructures. Recently interacting and multiplex networks are attracting
great attention because their dynamics and robustness cannot be understood
without treating at the same time several networks. Here we present maximal
entropy ensembles of spatial multiplex and spatial interacting networks that
can be used in order to model spatial multilayer network structures and to
build null models of real datasets. We show that spatial multiplex naturally
develop a significant overlap of the links, a noticeable property of many
multiplexes that can affect significantly the dynamics taking place on them.
Additionally, we characterize ensembles of spatial interacting networks and we
analyse the structure of interacting airport and railway networks in India,
showing the effect of space in determining the link probability.
|
1309.7031 | Nicola Perra | Suyu Liu, Nicola Perra, Marton Karsai, Alessandro Vespignani | Controlling Contagion Processes in Time-Varying Networks | null | null | 10.1103/PhysRevLett.112.118702 | null | physics.soc-ph cs.SI q-bio.PE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The vast majority of strategies aimed at controlling contagion processes on
networks considers the connectivity pattern of the system as either quenched or
annealed. However, in the real world many networks are highly dynamical and
evolve in time concurrently to the contagion process. Here, we derive an
analytical framework for the study of control strategies specifically devised
for time-varying networks. We consider the removal/immunization of individual
nodes according the their activity in the network and develop a block variable
mean-field approach that allows the derivation of the equations describing the
evolution of the contagion process concurrently to the network dynamic. We
derive the critical immunization threshold and assess the effectiveness of the
control strategies. Finally, we validate the theoretical picture by simulating
numerically the information spreading process and control strategies in both
synthetic networks and a large-scale, real-world mobile telephone call dataset
| [
{
"version": "v1",
"created": "Thu, 26 Sep 2013 19:50:15 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Liu",
"Suyu",
""
],
[
"Perra",
"Nicola",
""
],
[
"Karsai",
"Marton",
""
],
[
"Vespignani",
"Alessandro",
""
]
] | TITLE: Controlling Contagion Processes in Time-Varying Networks
ABSTRACT: The vast majority of strategies aimed at controlling contagion processes on
networks considers the connectivity pattern of the system as either quenched or
annealed. However, in the real world many networks are highly dynamical and
evolve in time concurrently to the contagion process. Here, we derive an
analytical framework for the study of control strategies specifically devised
for time-varying networks. We consider the removal/immunization of individual
nodes according the their activity in the network and develop a block variable
mean-field approach that allows the derivation of the equations describing the
evolution of the contagion process concurrently to the network dynamic. We
derive the critical immunization threshold and assess the effectiveness of the
control strategies. Finally, we validate the theoretical picture by simulating
numerically the information spreading process and control strategies in both
synthetic networks and a large-scale, real-world mobile telephone call dataset
|
1310.2632 | Philip Schniter | Jason T. Parker, Philip Schniter, and Volkan Cevher | Bilinear Generalized Approximate Message Passing | null | null | 10.1109/TSP.2014.2357776 | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We extend the generalized approximate message passing (G-AMP) approach,
originally proposed for high-dimensional generalized-linear regression in the
context of compressive sensing, to the generalized-bilinear case, which enables
its application to matrix completion, robust PCA, dictionary learning, and
related matrix-factorization problems. In the first part of the paper, we
derive our Bilinear G-AMP (BiG-AMP) algorithm as an approximation of the
sum-product belief propagation algorithm in the high-dimensional limit, where
central-limit theorem arguments and Taylor-series approximations apply, and
under the assumption of statistically independent matrix entries with known
priors. In addition, we propose an adaptive damping mechanism that aids
convergence under finite problem sizes, an expectation-maximization (EM)-based
method to automatically tune the parameters of the assumed priors, and two
rank-selection strategies. In the second part of the paper, we discuss the
specializations of EM-BiG-AMP to the problems of matrix completion, robust PCA,
and dictionary learning, and present the results of an extensive empirical
study comparing EM-BiG-AMP to state-of-the-art algorithms on each problem. Our
numerical results, using both synthetic and real-world datasets, demonstrate
that EM-BiG-AMP yields excellent reconstruction accuracy (often best in class)
while maintaining competitive runtimes and avoiding the need to tune
algorithmic parameters.
| [
{
"version": "v1",
"created": "Wed, 9 Oct 2013 21:08:40 GMT"
},
{
"version": "v2",
"created": "Fri, 1 Nov 2013 16:45:09 GMT"
},
{
"version": "v3",
"created": "Thu, 5 Jun 2014 14:32:06 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Parker",
"Jason T.",
""
],
[
"Schniter",
"Philip",
""
],
[
"Cevher",
"Volkan",
""
]
] | TITLE: Bilinear Generalized Approximate Message Passing
ABSTRACT: We extend the generalized approximate message passing (G-AMP) approach,
originally proposed for high-dimensional generalized-linear regression in the
context of compressive sensing, to the generalized-bilinear case, which enables
its application to matrix completion, robust PCA, dictionary learning, and
related matrix-factorization problems. In the first part of the paper, we
derive our Bilinear G-AMP (BiG-AMP) algorithm as an approximation of the
sum-product belief propagation algorithm in the high-dimensional limit, where
central-limit theorem arguments and Taylor-series approximations apply, and
under the assumption of statistically independent matrix entries with known
priors. In addition, we propose an adaptive damping mechanism that aids
convergence under finite problem sizes, an expectation-maximization (EM)-based
method to automatically tune the parameters of the assumed priors, and two
rank-selection strategies. In the second part of the paper, we discuss the
specializations of EM-BiG-AMP to the problems of matrix completion, robust PCA,
and dictionary learning, and present the results of an extensive empirical
study comparing EM-BiG-AMP to state-of-the-art algorithms on each problem. Our
numerical results, using both synthetic and real-world datasets, demonstrate
that EM-BiG-AMP yields excellent reconstruction accuracy (often best in class)
while maintaining competitive runtimes and avoiding the need to tune
algorithmic parameters.
|
1311.1753 | Rolf Andreassen | R. Andreassen, B. T. Meadows, M. de Silva, M. D. Sokoloff, and K.
Tomko | GooFit: A library for massively parallelising maximum-likelihood fits | Presented at the CHEP 2013 conference | null | 10.1088/1742-6596/513/5/052003 | null | cs.DC cs.MS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fitting complicated models to large datasets is a bottleneck of many
analyses. We present GooFit, a library and tool for constructing
arbitrarily-complex probability density functions (PDFs) to be evaluated on
nVidia GPUs or on multicore CPUs using OpenMP. The massive parallelisation of
dividing up event calculations between hundreds of processors can achieve
speedups of factors 200-300 in real-world problems.
| [
{
"version": "v1",
"created": "Thu, 7 Nov 2013 17:18:42 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Andreassen",
"R.",
""
],
[
"Meadows",
"B. T.",
""
],
[
"de Silva",
"M.",
""
],
[
"Sokoloff",
"M. D.",
""
],
[
"Tomko",
"K.",
""
]
] | TITLE: GooFit: A library for massively parallelising maximum-likelihood fits
ABSTRACT: Fitting complicated models to large datasets is a bottleneck of many
analyses. We present GooFit, a library and tool for constructing
arbitrarily-complex probability density functions (PDFs) to be evaluated on
nVidia GPUs or on multicore CPUs using OpenMP. The massive parallelisation of
dividing up event calculations between hundreds of processors can achieve
speedups of factors 200-300 in real-world problems.
|
1311.2911 | Kevin Kung | Kevin S. Kung, Kael Greco, Stanislav Sobolevsky, and Carlo Ratti | Exploring universal patterns in human home-work commuting from mobile
phone data | null | Kung KS, Greco K, Sobolevsky S, Ratti C (2014) Exploring Universal
Patterns in Human Home-Work Commuting from Mobile Phone Data. PLoS ONE 9(6):
e96180 | 10.1371/journal.pone.0096180 | null | cs.SI cs.CY physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Home-work commuting has always attracted significant research attention
because of its impact on human mobility. One of the key assumptions in this
domain of study is the universal uniformity of commute times. However, a true
comparison of commute patterns has often been hindered by the intrinsic
differences in data collection methods, which make observation from different
countries potentially biased and unreliable. In the present work, we approach
this problem through the use of mobile phone call detail records (CDRs), which
offers a consistent method for investigating mobility patterns in wholly
different parts of the world. We apply our analysis to a broad range of
datasets, at both the country and city scale. Additionally, we compare these
results with those obtained from vehicle GPS traces in Milan. While different
regions have some unique commute time characteristics, we show that the
home-work time distributions and average values within a single region are
indeed largely independent of commute distance or country (Portugal, Ivory
Coast, and Boston)--despite substantial spatial and infrastructural
differences. Furthermore, a comparative analysis demonstrates that such
distance-independence holds true only if we consider multimodal commute
behaviors--as consistent with previous studies. In car-only (Milan GPS traces)
and car-heavy (Saudi Arabia) commute datasets, we see that commute time is
indeed influenced by commute distance.
| [
{
"version": "v1",
"created": "Tue, 12 Nov 2013 20:29:14 GMT"
},
{
"version": "v2",
"created": "Wed, 24 Sep 2014 20:46:56 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Kung",
"Kevin S.",
""
],
[
"Greco",
"Kael",
""
],
[
"Sobolevsky",
"Stanislav",
""
],
[
"Ratti",
"Carlo",
""
]
] | TITLE: Exploring universal patterns in human home-work commuting from mobile
phone data
ABSTRACT: Home-work commuting has always attracted significant research attention
because of its impact on human mobility. One of the key assumptions in this
domain of study is the universal uniformity of commute times. However, a true
comparison of commute patterns has often been hindered by the intrinsic
differences in data collection methods, which make observation from different
countries potentially biased and unreliable. In the present work, we approach
this problem through the use of mobile phone call detail records (CDRs), which
offers a consistent method for investigating mobility patterns in wholly
different parts of the world. We apply our analysis to a broad range of
datasets, at both the country and city scale. Additionally, we compare these
results with those obtained from vehicle GPS traces in Milan. While different
regions have some unique commute time characteristics, we show that the
home-work time distributions and average values within a single region are
indeed largely independent of commute distance or country (Portugal, Ivory
Coast, and Boston)--despite substantial spatial and infrastructural
differences. Furthermore, a comparative analysis demonstrates that such
distance-independence holds true only if we consider multimodal commute
behaviors--as consistent with previous studies. In car-only (Milan GPS traces)
and car-heavy (Saudi Arabia) commute datasets, we see that commute time is
indeed influenced by commute distance.
|
1311.3928 | Stefan Kluth | Stefan Kluth | HS06 Benchmark for an ARM Server | three pages, two figures | null | 10.1088/1742-6596/513/6/062025 | MPP-2013-295 | physics.comp-ph cs.PF hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We benchmarked an ARM cortex-A9 based server system with a four-core CPU
running at 1.1 GHz. The system used Ubuntu 12.04 as operating system and the
HEPSPEC 2006 (HS06) benchmarking suite was compiled natively with gcc-4.4 on
the system. The benchmark was run for various settings of the relevant gcc
compiler options. We did not find significant influence from the compiler
options on the benchmark result. The final HS06 benchmark result is 10.4.
| [
{
"version": "v1",
"created": "Fri, 15 Nov 2013 17:17:21 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Kluth",
"Stefan",
""
]
] | TITLE: HS06 Benchmark for an ARM Server
ABSTRACT: We benchmarked an ARM cortex-A9 based server system with a four-core CPU
running at 1.1 GHz. The system used Ubuntu 12.04 as operating system and the
HEPSPEC 2006 (HS06) benchmarking suite was compiled natively with gcc-4.4 on
the system. The benchmark was run for various settings of the relevant gcc
compiler options. We did not find significant influence from the compiler
options on the benchmark result. The final HS06 benchmark result is 10.4.
|
1407.7390 | Jos\'e Ram\'on Padilla-L\'opez | Jos\'e Ram\'on Padilla-L\'opez and Alexandros Andr\'e Chaaraoui and
Francisco Fl\'orez-Revuelta | A discussion on the validation tests employed to compare human action
recognition methods using the MSR Action3D dataset | 16 pages and 7 tables | null | null | hdl:10045/39889 | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper aims to determine which is the best human action recognition
method based on features extracted from RGB-D devices, such as the Microsoft
Kinect. A review of all the papers that make reference to MSR Action3D, the
most used dataset that includes depth information acquired from a RGB-D device,
has been performed. We found that the validation method used by each work
differs from the others. So, a direct comparison among works cannot be made.
However, almost all the works present their results comparing them without
taking into account this issue. Therefore, we present different rankings
according to the methodology used for the validation in orden to clarify the
existing confusion.
| [
{
"version": "v1",
"created": "Mon, 28 Jul 2014 11:59:30 GMT"
},
{
"version": "v2",
"created": "Mon, 12 Jan 2015 11:30:40 GMT"
},
{
"version": "v3",
"created": "Tue, 16 Jun 2015 19:57:45 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Padilla-López",
"José Ramón",
""
],
[
"Chaaraoui",
"Alexandros André",
""
],
[
"Flórez-Revuelta",
"Francisco",
""
]
] | TITLE: A discussion on the validation tests employed to compare human action
recognition methods using the MSR Action3D dataset
ABSTRACT: This paper aims to determine which is the best human action recognition
method based on features extracted from RGB-D devices, such as the Microsoft
Kinect. A review of all the papers that make reference to MSR Action3D, the
most used dataset that includes depth information acquired from a RGB-D device,
has been performed. We found that the validation method used by each work
differs from the others. So, a direct comparison among works cannot be made.
However, almost all the works present their results comparing them without
taking into account this issue. Therefore, we present different rankings
according to the methodology used for the validation in orden to clarify the
existing confusion.
|
1506.01071 | Aleksey Buzmakov | Aleksey Buzmakov and Sergei O. Kuznetsov and Amedeo Napoli | Fast Generation of Best Interval Patterns for Nonmonotonic Constraints | 18 pages; 2 figures; 2 tables; 1 algorithm; PKDD 2015 Conference
Scientific Track | null | null | null | cs.AI cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In pattern mining, the main challenge is the exponential explosion of the set
of patterns. Typically, to solve this problem, a constraint for pattern
selection is introduced. One of the first constraints proposed in pattern
mining is support (frequency) of a pattern in a dataset. Frequency is an
anti-monotonic function, i.e., given an infrequent pattern, all its
superpatterns are not frequent. However, many other constraints for pattern
selection are not (anti-)monotonic, which makes it difficult to generate
patterns satisfying these constraints. In this paper we introduce the notion of
projection-antimonotonicity and $\theta$-$\Sigma\o\phi\iota\alpha$ algorithm
that allows efficient generation of the best patterns for some nonmonotonic
constraints. In this paper we consider stability and $\Delta$-measure, which
are nonmonotonic constraints, and apply them to interval tuple datasets. In the
experiments, we compute best interval tuple patterns w.r.t. these measures and
show the advantage of our approach over postfiltering approaches.
KEYWORDS: Pattern mining, nonmonotonic constraints, interval tuple data
| [
{
"version": "v1",
"created": "Tue, 2 Jun 2015 21:32:14 GMT"
},
{
"version": "v2",
"created": "Tue, 16 Jun 2015 15:31:19 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Buzmakov",
"Aleksey",
""
],
[
"Kuznetsov",
"Sergei O.",
""
],
[
"Napoli",
"Amedeo",
""
]
] | TITLE: Fast Generation of Best Interval Patterns for Nonmonotonic Constraints
ABSTRACT: In pattern mining, the main challenge is the exponential explosion of the set
of patterns. Typically, to solve this problem, a constraint for pattern
selection is introduced. One of the first constraints proposed in pattern
mining is support (frequency) of a pattern in a dataset. Frequency is an
anti-monotonic function, i.e., given an infrequent pattern, all its
superpatterns are not frequent. However, many other constraints for pattern
selection are not (anti-)monotonic, which makes it difficult to generate
patterns satisfying these constraints. In this paper we introduce the notion of
projection-antimonotonicity and $\theta$-$\Sigma\o\phi\iota\alpha$ algorithm
that allows efficient generation of the best patterns for some nonmonotonic
constraints. In this paper we consider stability and $\Delta$-measure, which
are nonmonotonic constraints, and apply them to interval tuple datasets. In the
experiments, we compute best interval tuple patterns w.r.t. these measures and
show the advantage of our approach over postfiltering approaches.
KEYWORDS: Pattern mining, nonmonotonic constraints, interval tuple data
|
1506.04757 | Julian McAuley | Julian McAuley, Christopher Targett, Qinfeng Shi, Anton van den Hengel | Image-based Recommendations on Styles and Substitutes | 11 pages, 10 figures, SIGIR 2015 | null | null | null | cs.CV cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Humans inevitably develop a sense of the relationships between objects, some
of which are based on their appearance. Some pairs of objects might be seen as
being alternatives to each other (such as two pairs of jeans), while others may
be seen as being complementary (such as a pair of jeans and a matching shirt).
This information guides many of the choices that people make, from buying
clothes to their interactions with each other. We seek here to model this human
sense of the relationships between objects based on their appearance. Our
approach is not based on fine-grained modeling of user annotations but rather
on capturing the largest dataset possible and developing a scalable method for
uncovering human notions of the visual relationships within. We cast this as a
network inference problem defined on graphs of related images, and provide a
large-scale dataset for the training and evaluation of the same. The system we
develop is capable of recommending which clothes and accessories will go well
together (and which will not), amongst a host of other applications.
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 20:01:49 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"McAuley",
"Julian",
""
],
[
"Targett",
"Christopher",
""
],
[
"Shi",
"Qinfeng",
""
],
[
"Hengel",
"Anton van den",
""
]
] | TITLE: Image-based Recommendations on Styles and Substitutes
ABSTRACT: Humans inevitably develop a sense of the relationships between objects, some
of which are based on their appearance. Some pairs of objects might be seen as
being alternatives to each other (such as two pairs of jeans), while others may
be seen as being complementary (such as a pair of jeans and a matching shirt).
This information guides many of the choices that people make, from buying
clothes to their interactions with each other. We seek here to model this human
sense of the relationships between objects based on their appearance. Our
approach is not based on fine-grained modeling of user annotations but rather
on capturing the largest dataset possible and developing a scalable method for
uncovering human notions of the visual relationships within. We cast this as a
network inference problem defined on graphs of related images, and provide a
large-scale dataset for the training and evaluation of the same. The system we
develop is capable of recommending which clothes and accessories will go well
together (and which will not), amongst a host of other applications.
|
1506.04776 | Jeff Heaton | Jeff Heaton | Encog: Library of Interchangeable Machine Learning Models for Java and
C# | null | null | null | null | cs.MS cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces the Encog library for Java and C#, a scalable,
adaptable, multiplatform machine learning framework that was 1st released in
2008. Encog allows a variety of machine learning models to be applied to
datasets using regression, classification, and clustering. Various supported
machine learning models can be used interchangeably with minimal recoding.
Encog uses efficient multithreaded code to reduce training time by exploiting
modern multicore processors. The current version of Encog can be downloaded
from http://www.encog.org.
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 21:20:06 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Heaton",
"Jeff",
""
]
] | TITLE: Encog: Library of Interchangeable Machine Learning Models for Java and
C#
ABSTRACT: This paper introduces the Encog library for Java and C#, a scalable,
adaptable, multiplatform machine learning framework that was 1st released in
2008. Encog allows a variety of machine learning models to be applied to
datasets using regression, classification, and clustering. Various supported
machine learning models can be used interchangeably with minimal recoding.
Encog uses efficient multithreaded code to reduce training time by exploiting
modern multicore processors. The current version of Encog can be downloaded
from http://www.encog.org.
|
1506.04803 | Afshin Rahimi | Afshin Rahimi, Duy Vu, Trevor Cohn, and Timothy Baldwin | Exploiting Text and Network Context for Geolocation of Social Media
Users | null | null | null | null | cs.CL cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Research on automatically geolocating social media users has conventionally
been based on the text content of posts from a given user or the social network
of the user, with very little crossover between the two, and no bench-marking
of the two approaches over compara- ble datasets. We bring the two threads of
research together in first proposing a text-based method based on adaptive
grids, followed by a hybrid network- and text-based method. Evaluating over
three Twitter datasets, we show that the empirical difference between text- and
network-based methods is not great, and that hybridisation of the two is
superior to the component methods, especially in contexts where the user graph
is not well connected. We achieve state-of-the-art results on all three
datasets.
| [
{
"version": "v1",
"created": "Tue, 16 Jun 2015 00:32:33 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Rahimi",
"Afshin",
""
],
[
"Vu",
"Duy",
""
],
[
"Cohn",
"Trevor",
""
],
[
"Baldwin",
"Timothy",
""
]
] | TITLE: Exploiting Text and Network Context for Geolocation of Social Media
Users
ABSTRACT: Research on automatically geolocating social media users has conventionally
been based on the text content of posts from a given user or the social network
of the user, with very little crossover between the two, and no bench-marking
of the two approaches over compara- ble datasets. We bring the two threads of
research together in first proposing a text-based method based on adaptive
grids, followed by a hybrid network- and text-based method. Evaluating over
three Twitter datasets, we show that the empirical difference between text- and
network-based methods is not great, and that hybridisation of the two is
superior to the component methods, especially in contexts where the user graph
is not well connected. We achieve state-of-the-art results on all three
datasets.
|
1506.04815 | Amit Chavan | Amit Chavan, Silu Huang, Amol Deshpande, Aaron Elmore, Samuel Madden
and Aditya Parameswaran | Towards a unified query language for provenance and versioning | Theory and Practice of Provenance, 2015 | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Organizations and teams collect and acquire data from various sources, such
as social interactions, financial transactions, sensor data, and genome
sequencers. Different teams in an organization as well as different data
scientists within a team are interested in extracting a variety of insights
which require combining and collaboratively analyzing datasets in diverse ways.
DataHub is a system that aims to provide robust version control and provenance
management for such a scenario. To be truly useful for collaborative data
science, one also needs the ability to specify queries and analysis tasks over
the versioning and the provenance information in a unified manner. In this
paper, we present an initial design of our query language, called VQuel, that
aims to support such unified querying over both types of information, as well
as the intermediate and final results of analyses. We also discuss some of the
key language design and implementation challenges moving forward.
| [
{
"version": "v1",
"created": "Tue, 16 Jun 2015 01:32:51 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Chavan",
"Amit",
""
],
[
"Huang",
"Silu",
""
],
[
"Deshpande",
"Amol",
""
],
[
"Elmore",
"Aaron",
""
],
[
"Madden",
"Samuel",
""
],
[
"Parameswaran",
"Aditya",
""
]
] | TITLE: Towards a unified query language for provenance and versioning
ABSTRACT: Organizations and teams collect and acquire data from various sources, such
as social interactions, financial transactions, sensor data, and genome
sequencers. Different teams in an organization as well as different data
scientists within a team are interested in extracting a variety of insights
which require combining and collaboratively analyzing datasets in diverse ways.
DataHub is a system that aims to provide robust version control and provenance
management for such a scenario. To be truly useful for collaborative data
science, one also needs the ability to specify queries and analysis tasks over
the versioning and the provenance information in a unified manner. In this
paper, we present an initial design of our query language, called VQuel, that
aims to support such unified querying over both types of information, as well
as the intermediate and final results of analyses. We also discuss some of the
key language design and implementation challenges moving forward.
|
1506.04945 | Mehul Bhatt | Carl Schultz and Mehul Bhatt | Spatial Symmetry Driven Pruning Strategies for Efficient Declarative
Spatial Reasoning | 22 pages. Accepted for publication at: COSIT 2015 - Conference on
Spatial Information Theory XII (COSIT), Santa Fe, New Mexico, USA ,October
2015 | null | null | null | cs.AI cs.LO cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Declarative spatial reasoning denotes the ability to (declaratively) specify
and solve real-world problems related to geometric and qualitative spatial
representation and reasoning within standard knowledge representation and
reasoning (KR) based methods (e.g., logic programming and derivatives). One
approach for encoding the semantics of spatial relations within a declarative
programming framework is by systems of polynomial constraints. However, solving
such constraints is computationally intractable in general (i.e. the theory of
real-closed fields).
We present a new algorithm, implemented within the declarative spatial
reasoning system CLP(QS), that drastically improves the performance of deciding
the consistency of spatial constraint graphs over conventional polynomial
encodings. We develop pruning strategies founded on spatial symmetries that
form equivalence classes (based on affine transformations) at the qualitative
spatial level. Moreover, pruning strategies are themselves formalised as
knowledge about the properties of space and spatial symmetries. We evaluate our
algorithm using a range of benchmarks in the class of contact problems, and
proofs in mereology and geometry. The empirical results show that CLP(QS) with
knowledge-based spatial pruning outperforms conventional polynomial encodings
by orders of magnitude, and can thus be applied to problems that are otherwise
unsolvable in practice.
| [
{
"version": "v1",
"created": "Tue, 16 Jun 2015 12:40:30 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Schultz",
"Carl",
""
],
[
"Bhatt",
"Mehul",
""
]
] | TITLE: Spatial Symmetry Driven Pruning Strategies for Efficient Declarative
Spatial Reasoning
ABSTRACT: Declarative spatial reasoning denotes the ability to (declaratively) specify
and solve real-world problems related to geometric and qualitative spatial
representation and reasoning within standard knowledge representation and
reasoning (KR) based methods (e.g., logic programming and derivatives). One
approach for encoding the semantics of spatial relations within a declarative
programming framework is by systems of polynomial constraints. However, solving
such constraints is computationally intractable in general (i.e. the theory of
real-closed fields).
We present a new algorithm, implemented within the declarative spatial
reasoning system CLP(QS), that drastically improves the performance of deciding
the consistency of spatial constraint graphs over conventional polynomial
encodings. We develop pruning strategies founded on spatial symmetries that
form equivalence classes (based on affine transformations) at the qualitative
spatial level. Moreover, pruning strategies are themselves formalised as
knowledge about the properties of space and spatial symmetries. We evaluate our
algorithm using a range of benchmarks in the class of contact problems, and
proofs in mereology and geometry. The empirical results show that CLP(QS) with
knowledge-based spatial pruning outperforms conventional polynomial encodings
by orders of magnitude, and can thus be applied to problems that are otherwise
unsolvable in practice.
|
1506.05001 | Liliana Lo Presti | Liliana Lo Presti and Marco La Cascia | Using Hankel Matrices for Dynamics-based Facial Emotion Recognition and
Pain Detection | in IEEE Proceedings of Workshop on Analysis and Modeling of Face and
Gestures (CVPRW 2015) | null | null | null | cs.CV cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a new approach to model the temporal dynamics of a
sequence of facial expressions. To this purpose, a sequence of Face Image
Descriptors (FID) is regarded as the output of a Linear Time Invariant (LTI)
system. The temporal dynamics of such sequence of descriptors are represented
by means of a Hankel matrix. The paper presents different strategies to compute
dynamics-based representation of a sequence of FID, and reports classification
accuracy values of the proposed representations within different standard
classification frameworks. The representations have been validated in two very
challenging application domains: emotion recognition and pain detection.
Experiments on two publicly available benchmarks and comparison with
state-of-the-art approaches demonstrate that the dynamics-based FID
representation attains competitive performance when off-the-shelf
classification tools are adopted.
| [
{
"version": "v1",
"created": "Tue, 16 Jun 2015 15:22:46 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Presti",
"Liliana Lo",
""
],
[
"La Cascia",
"Marco",
""
]
] | TITLE: Using Hankel Matrices for Dynamics-based Facial Emotion Recognition and
Pain Detection
ABSTRACT: This paper proposes a new approach to model the temporal dynamics of a
sequence of facial expressions. To this purpose, a sequence of Face Image
Descriptors (FID) is regarded as the output of a Linear Time Invariant (LTI)
system. The temporal dynamics of such sequence of descriptors are represented
by means of a Hankel matrix. The paper presents different strategies to compute
dynamics-based representation of a sequence of FID, and reports classification
accuracy values of the proposed representations within different standard
classification frameworks. The representations have been validated in two very
challenging application domains: emotion recognition and pain detection.
Experiments on two publicly available benchmarks and comparison with
state-of-the-art approaches demonstrate that the dynamics-based FID
representation attains competitive performance when off-the-shelf
classification tools are adopted.
|
1506.05101 | Dhruba Bhattacharyya | Hirak Kashyap, Hasin Afzal Ahmed, Nazrul Hoque, Swarup Roy and Dhruba
Kumar Bhattacharyya | Big Data Analytics in Bioinformatics: A Machine Learning Perspective | 20 pages survey paper on Big data analytics in Bioinformatics | null | null | null | cs.CE cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bioinformatics research is characterized by voluminous and incremental
datasets and complex data analytics methods. The machine learning methods used
in bioinformatics are iterative and parallel. These methods can be scaled to
handle big data using the distributed and parallel computing technologies.
Usually big data tools perform computation in batch-mode and are not
optimized for iterative processing and high data dependency among operations.
In the recent years, parallel, incremental, and multi-view machine learning
algorithms have been proposed. Similarly, graph-based architectures and
in-memory big data tools have been developed to minimize I/O cost and optimize
iterative processing.
However, there lack standard big data architectures and tools for many
important bioinformatics problems, such as fast construction of co-expression
and regulatory networks and salient module identification, detection of
complexes over growing protein-protein interaction data, fast analysis of
massive DNA, RNA, and protein sequence data, and fast querying on incremental
and heterogeneous disease networks. This paper addresses the issues and
challenges posed by several big data problems in bioinformatics, and gives an
overview of the state of the art and the future research opportunities.
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 11:32:00 GMT"
}
] | 2015-06-17T00:00:00 | [
[
"Kashyap",
"Hirak",
""
],
[
"Ahmed",
"Hasin Afzal",
""
],
[
"Hoque",
"Nazrul",
""
],
[
"Roy",
"Swarup",
""
],
[
"Bhattacharyya",
"Dhruba Kumar",
""
]
] | TITLE: Big Data Analytics in Bioinformatics: A Machine Learning Perspective
ABSTRACT: Bioinformatics research is characterized by voluminous and incremental
datasets and complex data analytics methods. The machine learning methods used
in bioinformatics are iterative and parallel. These methods can be scaled to
handle big data using the distributed and parallel computing technologies.
Usually big data tools perform computation in batch-mode and are not
optimized for iterative processing and high data dependency among operations.
In the recent years, parallel, incremental, and multi-view machine learning
algorithms have been proposed. Similarly, graph-based architectures and
in-memory big data tools have been developed to minimize I/O cost and optimize
iterative processing.
However, there lack standard big data architectures and tools for many
important bioinformatics problems, such as fast construction of co-expression
and regulatory networks and salient module identification, detection of
complexes over growing protein-protein interaction data, fast analysis of
massive DNA, RNA, and protein sequence data, and fast querying on incremental
and heterogeneous disease networks. This paper addresses the issues and
challenges posed by several big data problems in bioinformatics, and gives an
overview of the state of the art and the future research opportunities.
|
1305.6581 | Maxime Viallet Dr | Maxime Viallet, Isabelle Baraffe, Rolf Walder | Comparison of different nonlinear solvers for 2D time-implicit stellar
hydrodynamics | Accepted for publication in A&A | null | 10.1051/0004-6361/201220725 | null | astro-ph.SR cs.NA math.NA physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time-implicit schemes are attractive since they allow numerical time steps
that are much larger than those permitted by the Courant-Friedrich-Lewy
criterion characterizing time-explicit methods. This advantage comes, however,
with a cost: the solution of a system of nonlinear equations is required at
each time step. In this work, the nonlinear system results from the
discretization of the hydrodynamical equations with the Crank-Nicholson scheme.
We compare the cost of different methods, based on Newton-Raphson iterations,
to solve this nonlinear system, and benchmark their performances against
time-explicit schemes. Since our general scientific objective is to model
stellar interiors, we use as test cases two realistic models for the convective
envelope of a red giant and a young Sun. Focusing on 2D simulations, we show
that the best performances are obtained with the quasi-Newton method proposed
by Broyden. Another important concern is the accuracy of implicit calculations.
Based on the study of an idealized problem, namely the advection of a single
vortex by a uniform flow, we show that there are two aspects: i) the nonlinear
solver has to be accurate enough to resolve the truncation error of the
numerical discretization, and ii) the time step has be small enough to resolve
the advection of eddies. We show that with these two conditions fulfilled, our
implicit methods exhibit similar accuracy to time-explicit schemes, which have
lower values for the time step and higher computational costs. Finally, we
discuss in the conclusion the applicability of these methods to fully implicit
3D calculations.
| [
{
"version": "v1",
"created": "Tue, 28 May 2013 18:44:17 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Viallet",
"Maxime",
""
],
[
"Baraffe",
"Isabelle",
""
],
[
"Walder",
"Rolf",
""
]
] | TITLE: Comparison of different nonlinear solvers for 2D time-implicit stellar
hydrodynamics
ABSTRACT: Time-implicit schemes are attractive since they allow numerical time steps
that are much larger than those permitted by the Courant-Friedrich-Lewy
criterion characterizing time-explicit methods. This advantage comes, however,
with a cost: the solution of a system of nonlinear equations is required at
each time step. In this work, the nonlinear system results from the
discretization of the hydrodynamical equations with the Crank-Nicholson scheme.
We compare the cost of different methods, based on Newton-Raphson iterations,
to solve this nonlinear system, and benchmark their performances against
time-explicit schemes. Since our general scientific objective is to model
stellar interiors, we use as test cases two realistic models for the convective
envelope of a red giant and a young Sun. Focusing on 2D simulations, we show
that the best performances are obtained with the quasi-Newton method proposed
by Broyden. Another important concern is the accuracy of implicit calculations.
Based on the study of an idealized problem, namely the advection of a single
vortex by a uniform flow, we show that there are two aspects: i) the nonlinear
solver has to be accurate enough to resolve the truncation error of the
numerical discretization, and ii) the time step has be small enough to resolve
the advection of eddies. We show that with these two conditions fulfilled, our
implicit methods exhibit similar accuracy to time-explicit schemes, which have
lower values for the time step and higher computational costs. Finally, we
discuss in the conclusion the applicability of these methods to fully implicit
3D calculations.
|
1412.6610 | Daniel Jiwoong Im | Daniel Jiwoong Im and Graham W. Taylor | Scoring and Classifying with Gated Auto-encoders | null | null | null | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Auto-encoders are perhaps the best-known non-probabilistic methods for
representation learning. They are conceptually simple and easy to train. Recent
theoretical work has shed light on their ability to capture manifold structure,
and drawn connections to density modelling. This has motivated researchers to
seek ways of auto-encoder scoring, which has furthered their use in
classification. Gated auto-encoders (GAEs) are an interesting and flexible
extension of auto-encoders which can learn transformations among different
images or pixel covariances within images. However, they have been much less
studied, theoretically or empirically. In this work, we apply a dynamical
systems view to GAEs, deriving a scoring function, and drawing connections to
Restricted Boltzmann Machines. On a set of deep learning benchmarks, we also
demonstrate their effectiveness for single and multi-label classification.
| [
{
"version": "v1",
"created": "Sat, 20 Dec 2014 05:46:05 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Feb 2015 18:05:21 GMT"
},
{
"version": "v3",
"created": "Mon, 2 Mar 2015 16:35:39 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Apr 2015 18:15:16 GMT"
},
{
"version": "v5",
"created": "Mon, 15 Jun 2015 00:25:47 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Im",
"Daniel Jiwoong",
""
],
[
"Taylor",
"Graham W.",
""
]
] | TITLE: Scoring and Classifying with Gated Auto-encoders
ABSTRACT: Auto-encoders are perhaps the best-known non-probabilistic methods for
representation learning. They are conceptually simple and easy to train. Recent
theoretical work has shed light on their ability to capture manifold structure,
and drawn connections to density modelling. This has motivated researchers to
seek ways of auto-encoder scoring, which has furthered their use in
classification. Gated auto-encoders (GAEs) are an interesting and flexible
extension of auto-encoders which can learn transformations among different
images or pixel covariances within images. However, they have been much less
studied, theoretically or empirically. In this work, we apply a dynamical
systems view to GAEs, deriving a scoring function, and drawing connections to
Restricted Boltzmann Machines. On a set of deep learning benchmarks, we also
demonstrate their effectiveness for single and multi-label classification.
|
1506.04257 | Matthew Malloy | Matthew L. Malloy, Scott Alfeld, Paul Barford | Contamination Estimation via Convex Relaxations | To appear, ISIT 2015 | null | null | null | cs.IT cs.LG math.IT math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Identifying anomalies and contamination in datasets is important in a wide
variety of settings. In this paper, we describe a new technique for estimating
contamination in large, discrete valued datasets. Our approach considers the
normal condition of the data to be specified by a model consisting of a set of
distributions. Our key contribution is in our approach to contamination
estimation. Specifically, we develop a technique that identifies the minimum
number of data points that must be discarded (i.e., the level of contamination)
from an empirical data set in order to match the model to within a specified
goodness-of-fit, controlled by a p-value. Appealing to results from large
deviations theory, we show a lower bound on the level of contamination is
obtained by solving a series of convex programs. Theoretical results guarantee
the bound converges at a rate of $O(\sqrt{\log(p)/p})$, where p is the size of
the empirical data set.
| [
{
"version": "v1",
"created": "Sat, 13 Jun 2015 11:51:52 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Malloy",
"Matthew L.",
""
],
[
"Alfeld",
"Scott",
""
],
[
"Barford",
"Paul",
""
]
] | TITLE: Contamination Estimation via Convex Relaxations
ABSTRACT: Identifying anomalies and contamination in datasets is important in a wide
variety of settings. In this paper, we describe a new technique for estimating
contamination in large, discrete valued datasets. Our approach considers the
normal condition of the data to be specified by a model consisting of a set of
distributions. Our key contribution is in our approach to contamination
estimation. Specifically, we develop a technique that identifies the minimum
number of data points that must be discarded (i.e., the level of contamination)
from an empirical data set in order to match the model to within a specified
goodness-of-fit, controlled by a p-value. Appealing to results from large
deviations theory, we show a lower bound on the level of contamination is
obtained by solving a series of convex programs. Theoretical results guarantee
the bound converges at a rate of $O(\sqrt{\log(p)/p})$, where p is the size of
the empirical data set.
|
1506.04359 | Yunwen Lei | Yunwen Lei and \"Ur\"un Dogan and Alexander Binder and Marius Kloft | Multi-class SVMs: From Tighter Data-Dependent Generalization Bounds to
Novel Algorithms | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper studies the generalization performance of multi-class
classification algorithms, for which we obtain, for the first time, a
data-dependent generalization error bound with a logarithmic dependence on the
class size, substantially improving the state-of-the-art linear dependence in
the existing data-dependent generalization analysis. The theoretical analysis
motivates us to introduce a new multi-class classification machine based on
$\ell_p$-norm regularization, where the parameter $p$ controls the complexity
of the corresponding bounds. We derive an efficient optimization algorithm
based on Fenchel duality theory. Benchmarks on several real-world datasets show
that the proposed algorithm can achieve significant accuracy gains over the
state of the art.
| [
{
"version": "v1",
"created": "Sun, 14 Jun 2015 08:07:23 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Lei",
"Yunwen",
""
],
[
"Dogan",
"Ürün",
""
],
[
"Binder",
"Alexander",
""
],
[
"Kloft",
"Marius",
""
]
] | TITLE: Multi-class SVMs: From Tighter Data-Dependent Generalization Bounds to
Novel Algorithms
ABSTRACT: This paper studies the generalization performance of multi-class
classification algorithms, for which we obtain, for the first time, a
data-dependent generalization error bound with a logarithmic dependence on the
class size, substantially improving the state-of-the-art linear dependence in
the existing data-dependent generalization analysis. The theoretical analysis
motivates us to introduce a new multi-class classification machine based on
$\ell_p$-norm regularization, where the parameter $p$ controls the complexity
of the corresponding bounds. We derive an efficient optimization algorithm
based on Fenchel duality theory. Benchmarks on several real-world datasets show
that the proposed algorithm can achieve significant accuracy gains over the
state of the art.
|
1506.04608 | Daja Abdul | Javairia Nazir, Mehreen Sirshar | Flow Segmentation in Dense Crowds | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A framework is proposed in this paper that is used to segment flow of dense
crowds. The flow field that is generated by the movement in the crowd is
treated just like an aperiodic dynamic system. On this flow field a grid of
particles is put over for particle advection by the use of a numerical
integration scheme. Then flow maps are generated which associates the initial
position of the particles with final position. The gradient of the flow maps
gives the amount of divergence of the neighboring particles. For forward
integration and analysis forward Finite time Lyapunov Exponent is calculated
and backward Finite time Lyapunov Exponent is also calculated it gives the
Lagrangian coherent structures of the flow in crowd. Lagrangian Coherent
Structures basically divides the flow in crowd into regions and these regions
have different dynamics. These regions are then used to get the boundary in the
different flow segments by using water shed algorithm. The experiment is
conducted on the crowd dataset of UCF (University of central Florida).
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 14:14:20 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Nazir",
"Javairia",
""
],
[
"Sirshar",
"Mehreen",
""
]
] | TITLE: Flow Segmentation in Dense Crowds
ABSTRACT: A framework is proposed in this paper that is used to segment flow of dense
crowds. The flow field that is generated by the movement in the crowd is
treated just like an aperiodic dynamic system. On this flow field a grid of
particles is put over for particle advection by the use of a numerical
integration scheme. Then flow maps are generated which associates the initial
position of the particles with final position. The gradient of the flow maps
gives the amount of divergence of the neighboring particles. For forward
integration and analysis forward Finite time Lyapunov Exponent is calculated
and backward Finite time Lyapunov Exponent is also calculated it gives the
Lagrangian coherent structures of the flow in crowd. Lagrangian Coherent
Structures basically divides the flow in crowd into regions and these regions
have different dynamics. These regions are then used to get the boundary in the
different flow segments by using water shed algorithm. The experiment is
conducted on the crowd dataset of UCF (University of central Florida).
|
1506.04693 | Vincent Labatut | G\"unce Orman (BIT Lab), Vincent Labatut (LIA), Marc Plantevit
(LIRIS), Jean-Fran\c{c}ois Boulicaut (LIRIS) | Interpreting communities based on the evolution of a dynamic attributed
network | null | Social Network Analysis and Mining Journal (SNAM), 2015, 5, pp.20.
\<http://link.springer.com/article/10.1007%2Fs13278-015-0262-4\>.
\<10.1007/s13278-015-0262-4\> | 10.1007/s13278-015-0262-4 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many methods have been proposed to detect communities, not only in plain, but
also in attributed, directed or even dynamic complex networks. From the
modeling point of view, to be of some utility, the community structure must be
characterized relatively to the properties of the studied system. However, most
of the existing works focus on the detection of communities, and only very few
try to tackle this interpretation problem. Moreover, the existing approaches
are limited either by the type of data they handle, or by the nature of the
results they output. In this work, we see the interpretation of communities as
a problem independent from the detection process, consisting in identifying the
most characteristic features of communities. We give a formal definition of
this problem and propose a method to solve it. To this aim, we first define a
sequence-based representation of networks, combining temporal information,
community structure, topological measures, and nodal attributes. We then
describe how to identify the most emerging sequential patterns of this dataset,
and use them to characterize the communities. We study the performance of our
method on artificially generated dynamic attributed networks. We also
empirically validate our framework on real-world systems: a DBLP network of
scientific collaborations, and a LastFM network of social and musical
interactions.
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 18:22:38 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Orman",
"Günce",
"",
"BIT Lab"
],
[
"Labatut",
"Vincent",
"",
"LIA"
],
[
"Plantevit",
"Marc",
"",
"LIRIS"
],
[
"Boulicaut",
"Jean-François",
"",
"LIRIS"
]
] | TITLE: Interpreting communities based on the evolution of a dynamic attributed
network
ABSTRACT: Many methods have been proposed to detect communities, not only in plain, but
also in attributed, directed or even dynamic complex networks. From the
modeling point of view, to be of some utility, the community structure must be
characterized relatively to the properties of the studied system. However, most
of the existing works focus on the detection of communities, and only very few
try to tackle this interpretation problem. Moreover, the existing approaches
are limited either by the type of data they handle, or by the nature of the
results they output. In this work, we see the interpretation of communities as
a problem independent from the detection process, consisting in identifying the
most characteristic features of communities. We give a formal definition of
this problem and propose a method to solve it. To this aim, we first define a
sequence-based representation of networks, combining temporal information,
community structure, topological measures, and nodal attributes. We then
describe how to identify the most emerging sequential patterns of this dataset,
and use them to characterize the communities. We study the performance of our
method on artificially generated dynamic attributed networks. We also
empirically validate our framework on real-world systems: a DBLP network of
scientific collaborations, and a LastFM network of social and musical
interactions.
|
1506.04720 | Siqi Nie | Siqi Nie, Qiang Ji | Latent Regression Bayesian Network for Data Representation | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep directed generative models have attracted much attention recently due to
their expressive representation power and the ability of ancestral sampling.
One major difficulty of learning directed models with many latent variables is
the intractable inference. To address this problem, most existing algorithms
make assumptions to render the latent variables independent of each other,
either by designing specific priors, or by approximating the true posterior
using a factorized distribution. We believe the correlations among latent
variables are crucial for faithful data representation. Driven by this idea, we
propose an inference method based on the conditional pseudo-likelihood that
preserves the dependencies among the latent variables. For learning, we propose
to employ the hard Expectation Maximization (EM) algorithm, which avoids the
intractability of the traditional EM by max-out instead of sum-out to compute
the data likelihood. Qualitative and quantitative evaluations of our model
against state of the art deep models on benchmark datasets demonstrate the
effectiveness of the proposed algorithm in data representation and
reconstruction.
| [
{
"version": "v1",
"created": "Mon, 15 Jun 2015 19:34:59 GMT"
}
] | 2015-06-16T00:00:00 | [
[
"Nie",
"Siqi",
""
],
[
"Ji",
"Qiang",
""
]
] | TITLE: Latent Regression Bayesian Network for Data Representation
ABSTRACT: Deep directed generative models have attracted much attention recently due to
their expressive representation power and the ability of ancestral sampling.
One major difficulty of learning directed models with many latent variables is
the intractable inference. To address this problem, most existing algorithms
make assumptions to render the latent variables independent of each other,
either by designing specific priors, or by approximating the true posterior
using a factorized distribution. We believe the correlations among latent
variables are crucial for faithful data representation. Driven by this idea, we
propose an inference method based on the conditional pseudo-likelihood that
preserves the dependencies among the latent variables. For learning, we propose
to employ the hard Expectation Maximization (EM) algorithm, which avoids the
intractability of the traditional EM by max-out instead of sum-out to compute
the data likelihood. Qualitative and quantitative evaluations of our model
against state of the art deep models on benchmark datasets demonstrate the
effectiveness of the proposed algorithm in data representation and
reconstruction.
|
1303.0452 | Zhengfeng Yang | Min Wu, Zhengfeng Yang and Wang Lin | Domain-of-Attraction Estimation for Uncertain Non-polynomial Systems | null | null | 10.1016/j.cnsns.2013.12.001 | null | cs.SC math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we consider the problem of computing estimates of the
domain-of-attraction for non-polynomial systems. A polynomial approximation
technique, based on multivariate polynomial interpolation and error analysis
for remaining functions, is applied to compute an uncertain polynomial system,
whose set of trajectories contains that of the original non-polynomial system.
Experiments on the benchmark non-polynomial systems show that our approach
gives better estimates of the domain-of-attraction.
| [
{
"version": "v1",
"created": "Sun, 3 Mar 2013 03:01:29 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Wu",
"Min",
""
],
[
"Yang",
"Zhengfeng",
""
],
[
"Lin",
"Wang",
""
]
] | TITLE: Domain-of-Attraction Estimation for Uncertain Non-polynomial Systems
ABSTRACT: In this paper, we consider the problem of computing estimates of the
domain-of-attraction for non-polynomial systems. A polynomial approximation
technique, based on multivariate polynomial interpolation and error analysis
for remaining functions, is applied to compute an uncertain polynomial system,
whose set of trajectories contains that of the original non-polynomial system.
Experiments on the benchmark non-polynomial systems show that our approach
gives better estimates of the domain-of-attraction.
|
1303.5855 | Zhong-Yuan Zhang | Zhong-Yuan Zhang and Yong Wang and Yong-Yeol Ahn | Overlapping Community Detection in Complex Networks using Symmetric
Binary Matrix Factorization | null | null | 10.1103/PhysRevE.87.062803 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Discovering overlapping community structures is a crucial step to
understanding the structure and dynamics of many networks. In this paper we
develop a symmetric binary matrix factorization model (SBMF) to identify
overlapping communities. Our model allows us not only to assign community
memberships explicitly to nodes, but also to distinguish outliers from
overlapping nodes. In addition, we propose a modified partition density to
evaluate the quality of community structures. We use this to determine the most
appropriate number of communities. We evaluate our methods using both synthetic
benchmarks and real world networks, demonstrating the effectiveness of our
approach.
| [
{
"version": "v1",
"created": "Sat, 23 Mar 2013 15:16:44 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Zhang",
"Zhong-Yuan",
""
],
[
"Wang",
"Yong",
""
],
[
"Ahn",
"Yong-Yeol",
""
]
] | TITLE: Overlapping Community Detection in Complex Networks using Symmetric
Binary Matrix Factorization
ABSTRACT: Discovering overlapping community structures is a crucial step to
understanding the structure and dynamics of many networks. In this paper we
develop a symmetric binary matrix factorization model (SBMF) to identify
overlapping communities. Our model allows us not only to assign community
memberships explicitly to nodes, but also to distinguish outliers from
overlapping nodes. In addition, we propose a modified partition density to
evaluate the quality of community structures. We use this to determine the most
appropriate number of communities. We evaluate our methods using both synthetic
benchmarks and real world networks, demonstrating the effectiveness of our
approach.
|
1303.6170 | Brandon Jones | Brandon Jones, Mark Campbell, Lang Tong | Maximum Likelihood Fusion of Stochastic Maps | 10 pages, 8 figures, submitted to IEEE Transactions on Signal
Processing on 24-March-2013 | null | 10.1109/TSP.2014.2304435 | null | stat.AP cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The fusion of independently obtained stochastic maps by collaborating mobile
agents is considered. The proposed approach includes two parts: matching of
stochastic maps and maximum likelihood alignment. In particular, an affine
invariant hypergraph is constructed for each stochastic map, and a bipartite
matching via a linear program is used to establish landmark correspondence
between stochastic maps. A maximum likelihood alignment procedure is proposed
to determine rotation and translation between common landmarks in order to
construct a global map within a common frame of reference. A main feature of
the proposed approach is its scalability with respect to the number of
landmarks: the matching step has polynomial complexity and the maximum
likelihood alignment is obtained in closed form. Experimental validation of the
proposed fusion approach is performed using the Victoria Park benchmark
dataset.
| [
{
"version": "v1",
"created": "Mon, 25 Mar 2013 15:34:26 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Jones",
"Brandon",
""
],
[
"Campbell",
"Mark",
""
],
[
"Tong",
"Lang",
""
]
] | TITLE: Maximum Likelihood Fusion of Stochastic Maps
ABSTRACT: The fusion of independently obtained stochastic maps by collaborating mobile
agents is considered. The proposed approach includes two parts: matching of
stochastic maps and maximum likelihood alignment. In particular, an affine
invariant hypergraph is constructed for each stochastic map, and a bipartite
matching via a linear program is used to establish landmark correspondence
between stochastic maps. A maximum likelihood alignment procedure is proposed
to determine rotation and translation between common landmarks in order to
construct a global map within a common frame of reference. A main feature of
the proposed approach is its scalability with respect to the number of
landmarks: the matching step has polynomial complexity and the maximum
likelihood alignment is obtained in closed form. Experimental validation of the
proposed fusion approach is performed using the Victoria Park benchmark
dataset.
|
1304.5302 | Satoshi Eguchi | Satoshi Eguchi | "Superluminal" FITS File Processing on Multiprocessors: Zero Time Endian
Conversion Technique | 25 pages, 9 figures, 12 tables, accepted for publication in PASP | null | 10.1086/671105 | null | astro-ph.IM cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The FITS is the standard file format in astronomy, and it has been extended
to agree with astronomical needs of the day. However, astronomical datasets
have been inflating year by year. In case of ALMA telescope, a ~ TB scale
4-dimensional data cube may be produced for one target. Considering that
typical Internet bandwidth is a few 10 MB/s at most, the original data cubes in
FITS format are hosted on a VO server, and the region which a user is
interested in should be cut out and transferred to the user (Eguchi et al.,
2012). The system will equip a very high-speed disk array to process a TB scale
data cube in a few 10 seconds, and disk I/O speed, endian conversion and data
processing one will be comparable. Hence to reduce the endian conversion time
is one of issues to realize our system. In this paper, I introduce a technique
named "just-in-time endian conversion", which delays the endian conversion for
each pixel just before it is really needed, to sweep out the endian conversion
time; by applying this method, the FITS processing speed increases 20% for
single threading, and 40% for multi-threading compared to CFITSIO. The speed-up
by the method tightly relates to modern CPU architecture to improve the
efficiency of instruction pipelines due to break of "causality", a programmed
instruction code sequence.
| [
{
"version": "v1",
"created": "Fri, 19 Apr 2013 03:29:36 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Eguchi",
"Satoshi",
""
]
] | TITLE: "Superluminal" FITS File Processing on Multiprocessors: Zero Time Endian
Conversion Technique
ABSTRACT: The FITS is the standard file format in astronomy, and it has been extended
to agree with astronomical needs of the day. However, astronomical datasets
have been inflating year by year. In case of ALMA telescope, a ~ TB scale
4-dimensional data cube may be produced for one target. Considering that
typical Internet bandwidth is a few 10 MB/s at most, the original data cubes in
FITS format are hosted on a VO server, and the region which a user is
interested in should be cut out and transferred to the user (Eguchi et al.,
2012). The system will equip a very high-speed disk array to process a TB scale
data cube in a few 10 seconds, and disk I/O speed, endian conversion and data
processing one will be comparable. Hence to reduce the endian conversion time
is one of issues to realize our system. In this paper, I introduce a technique
named "just-in-time endian conversion", which delays the endian conversion for
each pixel just before it is really needed, to sweep out the endian conversion
time; by applying this method, the FITS processing speed increases 20% for
single threading, and 40% for multi-threading compared to CFITSIO. The speed-up
by the method tightly relates to modern CPU architecture to improve the
efficiency of instruction pipelines due to break of "causality", a programmed
instruction code sequence.
|
1305.3532 | Alain Barrat | Alain Barrat, Ciro Cattuto | Temporal networks of face-to-face human interactions | Chapter of the book "Temporal Networks", Springer, 2013. Series:
Understanding Complex Systems. Holme, Petter; Saram\"aki, Jari (Eds.) | null | 10.1007/978-3-642-36461-7_10 | null | physics.soc-ph cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ever increasing adoption of mobile technologies and ubiquitous services
allows to sense human behavior at unprecedented levels of details and scale.
Wearable sensors are opening up a new window on human mobility and proximity at
the finest resolution of face-to-face proximity. As a consequence, empirical
data describing social and behavioral networks are acquiring a longitudinal
dimension that brings forth new challenges for analysis and modeling. Here we
review recent work on the representation and analysis of temporal networks of
face-to-face human proximity, based on large-scale datasets collected in the
context of the SocioPatterns collaboration. We show that the raw behavioral
data can be studied at various levels of coarse-graining, which turn out to be
complementary to one another, with each level exposing different features of
the underlying system. We briefly review a generative model of temporal contact
networks that reproduces some statistical observables. Then, we shift our focus
from surface statistical features to dynamical processes on empirical temporal
networks. We discuss how simple dynamical processes can be used as probes to
expose important features of the interaction patterns, such as burstiness and
causal constraints. We show that simulating dynamical processes on empirical
temporal networks can unveil differences between datasets that would otherwise
look statistically similar. Moreover, we argue that, due to the temporal
heterogeneity of human dynamics, in order to investigate the temporal
properties of spreading processes it may be necessary to abandon the notion of
wall-clock time in favour of an intrinsic notion of time for each individual
node, defined in terms of its activity level. We conclude highlighting several
open research questions raised by the nature of the data at hand.
| [
{
"version": "v1",
"created": "Wed, 15 May 2013 16:18:24 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Barrat",
"Alain",
""
],
[
"Cattuto",
"Ciro",
""
]
] | TITLE: Temporal networks of face-to-face human interactions
ABSTRACT: The ever increasing adoption of mobile technologies and ubiquitous services
allows to sense human behavior at unprecedented levels of details and scale.
Wearable sensors are opening up a new window on human mobility and proximity at
the finest resolution of face-to-face proximity. As a consequence, empirical
data describing social and behavioral networks are acquiring a longitudinal
dimension that brings forth new challenges for analysis and modeling. Here we
review recent work on the representation and analysis of temporal networks of
face-to-face human proximity, based on large-scale datasets collected in the
context of the SocioPatterns collaboration. We show that the raw behavioral
data can be studied at various levels of coarse-graining, which turn out to be
complementary to one another, with each level exposing different features of
the underlying system. We briefly review a generative model of temporal contact
networks that reproduces some statistical observables. Then, we shift our focus
from surface statistical features to dynamical processes on empirical temporal
networks. We discuss how simple dynamical processes can be used as probes to
expose important features of the interaction patterns, such as burstiness and
causal constraints. We show that simulating dynamical processes on empirical
temporal networks can unveil differences between datasets that would otherwise
look statistically similar. Moreover, we argue that, due to the temporal
heterogeneity of human dynamics, in order to investigate the temporal
properties of spreading processes it may be necessary to abandon the notion of
wall-clock time in favour of an intrinsic notion of time for each individual
node, defined in terms of its activity level. We conclude highlighting several
open research questions raised by the nature of the data at hand.
|
1412.3421 | Juan Eugenio Iglesias | Juan Eugenio Iglesias and Mert Rory Sabuncu | Multi-Atlas Segmentation of Biomedical Images: A Survey | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-atlas segmentation (MAS), first introduced and popularized by the
pioneering work of Rohlfing, Brandt, Menzel and Maurer Jr (2004), Klein, Mensh,
Ghosh, Tourville and Hirsch (2005), and Heckemann, Hajnal, Aljabar, Rueckert
and Hammers (2006), is becoming one of the most widely-used and successful
image segmentation techniques in biomedical applications. By manipulating and
utilizing the entire dataset of "atlases" (training images that have been
previously labeled, e.g., manually by an expert), rather than some model-based
average representation, MAS has the flexibility to better capture anatomical
variation, thus offering superior segmentation accuracy. This benefit, however,
typically comes at a high computational cost. Recent advancements in computer
hardware and image processing software have been instrumental in addressing
this challenge and facilitated the wide adoption of MAS. Today, MAS has come a
long way and the approach includes a wide array of sophisticated algorithms
that employ ideas from machine learning, probabilistic modeling, optimization,
and computer vision, among other fields. This paper presents a survey of
published MAS algorithms and studies that have applied these methods to various
biomedical problems. In writing this survey, we have three distinct aims. Our
primary goal is to document how MAS was originally conceived, later evolved,
and now relates to alternative methods. Second, this paper is intended to be a
detailed reference of past research activity in MAS, which now spans over a
decade (2003 - 2014) and entails novel methodological developments and
application-specific solutions. Finally, our goal is to also present a
perspective on the future of MAS, which, we believe, will be one of the
dominant approaches in biomedical image segmentation.
| [
{
"version": "v1",
"created": "Wed, 10 Dec 2014 19:28:09 GMT"
},
{
"version": "v2",
"created": "Fri, 12 Jun 2015 14:35:30 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Iglesias",
"Juan Eugenio",
""
],
[
"Sabuncu",
"Mert Rory",
""
]
] | TITLE: Multi-Atlas Segmentation of Biomedical Images: A Survey
ABSTRACT: Multi-atlas segmentation (MAS), first introduced and popularized by the
pioneering work of Rohlfing, Brandt, Menzel and Maurer Jr (2004), Klein, Mensh,
Ghosh, Tourville and Hirsch (2005), and Heckemann, Hajnal, Aljabar, Rueckert
and Hammers (2006), is becoming one of the most widely-used and successful
image segmentation techniques in biomedical applications. By manipulating and
utilizing the entire dataset of "atlases" (training images that have been
previously labeled, e.g., manually by an expert), rather than some model-based
average representation, MAS has the flexibility to better capture anatomical
variation, thus offering superior segmentation accuracy. This benefit, however,
typically comes at a high computational cost. Recent advancements in computer
hardware and image processing software have been instrumental in addressing
this challenge and facilitated the wide adoption of MAS. Today, MAS has come a
long way and the approach includes a wide array of sophisticated algorithms
that employ ideas from machine learning, probabilistic modeling, optimization,
and computer vision, among other fields. This paper presents a survey of
published MAS algorithms and studies that have applied these methods to various
biomedical problems. In writing this survey, we have three distinct aims. Our
primary goal is to document how MAS was originally conceived, later evolved,
and now relates to alternative methods. Second, this paper is intended to be a
detailed reference of past research activity in MAS, which now spans over a
decade (2003 - 2014) and entails novel methodological developments and
application-specific solutions. Finally, our goal is to also present a
perspective on the future of MAS, which, we believe, will be one of the
dominant approaches in biomedical image segmentation.
|
1501.06625 | Jan Verschelde | Jan Verschelde and Xiangcheng Yu | Accelerating Polynomial Homotopy Continuation on a Graphics Processing
Unit with Double Double and Quad Double Arithmetic | Accepted for publication in the Proceedings of the 7th International
Workshop on Parallel Symbolic Computation (PASCO 2015) | null | null | null | cs.MS cs.DC cs.NA math.AG math.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Numerical continuation methods track a solution path defined by a homotopy.
The systems we consider are defined by polynomials in several variables with
complex coefficients. For larger dimensions and degrees, the numerical
conditioning worsens and hardware double precision becomes often insufficient
to reach the end of the solution path. With double double and quad double
arithmetic, we can solve larger problems that we could not solve with hardware
double arithmetic, but at a higher computational cost. This cost overhead can
be compensated by acceleration on a Graphics Processing Unit (GPU). We describe
our implementation and report on computational results on benchmark polynomial
systems.
| [
{
"version": "v1",
"created": "Mon, 26 Jan 2015 23:54:20 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Apr 2015 15:55:11 GMT"
},
{
"version": "v3",
"created": "Fri, 12 Jun 2015 15:08:55 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Verschelde",
"Jan",
""
],
[
"Yu",
"Xiangcheng",
""
]
] | TITLE: Accelerating Polynomial Homotopy Continuation on a Graphics Processing
Unit with Double Double and Quad Double Arithmetic
ABSTRACT: Numerical continuation methods track a solution path defined by a homotopy.
The systems we consider are defined by polynomials in several variables with
complex coefficients. For larger dimensions and degrees, the numerical
conditioning worsens and hardware double precision becomes often insufficient
to reach the end of the solution path. With double double and quad double
arithmetic, we can solve larger problems that we could not solve with hardware
double arithmetic, but at a higher computational cost. This cost overhead can
be compensated by acceleration on a Graphics Processing Unit (GPU). We describe
our implementation and report on computational results on benchmark polynomial
systems.
|
1503.01212 | Alexander Rakhlin | Alexander Rakhlin, Karthik Sridharan | Hierarchies of Relaxations for Online Prediction Problems with Evolving
Constraints | null | null | null | null | cs.LG cs.DS stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study online prediction where regret of the algorithm is measured against
a benchmark defined via evolving constraints. This framework captures online
prediction on graphs, as well as other prediction problems with combinatorial
structure. A key aspect here is that finding the optimal benchmark predictor
(even in hindsight, given all the data) might be computationally hard due to
the combinatorial nature of the constraints. Despite this, we provide
polynomial-time \emph{prediction} algorithms that achieve low regret against
combinatorial benchmark sets. We do so by building improper learning algorithms
based on two ideas that work together. The first is to alleviate part of the
computational burden through random playout, and the second is to employ
Lasserre semidefinite hierarchies to approximate the resulting integer program.
Interestingly, for our prediction algorithms, we only need to compute the
values of the semidefinite programs and not the rounded solutions. However, the
integrality gap for Lasserre hierarchy \emph{does} enter the generic regret
bound in terms of Rademacher complexity of the benchmark set. This establishes
a trade-off between the computation time and the regret bound of the algorithm.
| [
{
"version": "v1",
"created": "Wed, 4 Mar 2015 04:05:35 GMT"
},
{
"version": "v2",
"created": "Fri, 12 Jun 2015 00:04:05 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Rakhlin",
"Alexander",
""
],
[
"Sridharan",
"Karthik",
""
]
] | TITLE: Hierarchies of Relaxations for Online Prediction Problems with Evolving
Constraints
ABSTRACT: We study online prediction where regret of the algorithm is measured against
a benchmark defined via evolving constraints. This framework captures online
prediction on graphs, as well as other prediction problems with combinatorial
structure. A key aspect here is that finding the optimal benchmark predictor
(even in hindsight, given all the data) might be computationally hard due to
the combinatorial nature of the constraints. Despite this, we provide
polynomial-time \emph{prediction} algorithms that achieve low regret against
combinatorial benchmark sets. We do so by building improper learning algorithms
based on two ideas that work together. The first is to alleviate part of the
computational burden through random playout, and the second is to employ
Lasserre semidefinite hierarchies to approximate the resulting integer program.
Interestingly, for our prediction algorithms, we only need to compute the
values of the semidefinite programs and not the rounded solutions. However, the
integrality gap for Lasserre hierarchy \emph{does} enter the generic regret
bound in terms of Rademacher complexity of the benchmark set. This establishes
a trade-off between the computation time and the regret bound of the algorithm.
|
1505.07194 | Peng Liu | Peng Liu and Saeed Gazor and Il-Min Kim and Dong In Kim | Noncoherent Relaying in Energy Harvesting Communication Systems | null | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by-nc-sa/3.0/ | In energy harvesting (EH) relay networks, the coherent communication requires
accurate estima- tion/tracking of the instantaneous channel state information
(CSI) which consumes extra power. As a remedy, we propose two noncoherent EH
relaying protocols based on the amplify-and-forward (AF) relaying, namely,
power splitting noncoherent AF (PS-NcAF) and time switching noncoherent AF
(TS-NcAF), which do not require any instantaneous CSI. We develop a noncoherent
framework of simultaneous wireless information and power transfer (SWIPT),
embracing PS-NcAF and TS-NcAF in a unified form. For arbitrary M-ary
noncoherent frequency-shift keying (FSK) and differential phase- shift keying
(DPSK), we derive maximum-likelihood detectors (MLDs) for PS-NcAF and TS-NcAF
in a unified form, which involves integral evaluations yet serves as the
optimum performance benchmark. To avoid expensive integral computations, we
propose a closed-form detector using the Gauss-Legendre approximation, which
achieves almost identical performance as the MLD but at substantially lower
complexity. These EH-based noncoherent detectors achieve full diversity in
Rayleigh fading. Numerical results demonstrate that our proposed PS-NcAF and
TS-NcAF may outperform the conventional grid- powered relay system under the
same total power constraint. Various insights which are useful for the design
of practical SWIPT relaying systems are obtained. Interestingly, PS-NcAF
outperforms TS-NcAF in the single-relay case, whereas TS-NcAF outperforms
PS-NcAF in the multi-relay case.
| [
{
"version": "v1",
"created": "Wed, 27 May 2015 05:56:15 GMT"
},
{
"version": "v2",
"created": "Fri, 12 Jun 2015 14:18:26 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Liu",
"Peng",
""
],
[
"Gazor",
"Saeed",
""
],
[
"Kim",
"Il-Min",
""
],
[
"Kim",
"Dong In",
""
]
] | TITLE: Noncoherent Relaying in Energy Harvesting Communication Systems
ABSTRACT: In energy harvesting (EH) relay networks, the coherent communication requires
accurate estima- tion/tracking of the instantaneous channel state information
(CSI) which consumes extra power. As a remedy, we propose two noncoherent EH
relaying protocols based on the amplify-and-forward (AF) relaying, namely,
power splitting noncoherent AF (PS-NcAF) and time switching noncoherent AF
(TS-NcAF), which do not require any instantaneous CSI. We develop a noncoherent
framework of simultaneous wireless information and power transfer (SWIPT),
embracing PS-NcAF and TS-NcAF in a unified form. For arbitrary M-ary
noncoherent frequency-shift keying (FSK) and differential phase- shift keying
(DPSK), we derive maximum-likelihood detectors (MLDs) for PS-NcAF and TS-NcAF
in a unified form, which involves integral evaluations yet serves as the
optimum performance benchmark. To avoid expensive integral computations, we
propose a closed-form detector using the Gauss-Legendre approximation, which
achieves almost identical performance as the MLD but at substantially lower
complexity. These EH-based noncoherent detectors achieve full diversity in
Rayleigh fading. Numerical results demonstrate that our proposed PS-NcAF and
TS-NcAF may outperform the conventional grid- powered relay system under the
same total power constraint. Various insights which are useful for the design
of practical SWIPT relaying systems are obtained. Interestingly, PS-NcAF
outperforms TS-NcAF in the single-relay case, whereas TS-NcAF outperforms
PS-NcAF in the multi-relay case.
|
1506.00815 | Yuhuang Hu | Yuhuang Hu, M.S. Ishwarya, Chu Kiong Loo | Classify Images with Conceptor Network | This paper has been withdrawn by the author due to a crucial sign
error in experiments | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This article demonstrates a new conceptor network based classifier in
classifying images. Mathematical descriptions and analysis are presented.
Various tests are experimented using three benchmark datasets: MNIST, CIFAR-10
and CIFAR-100. The experiments displayed that conceptor network can offer
superior results and flexible configurations than conventional classifiers such
as Softmax Regression and Support Vector Machine (SVM).
| [
{
"version": "v1",
"created": "Tue, 2 Jun 2015 09:49:45 GMT"
},
{
"version": "v2",
"created": "Wed, 3 Jun 2015 13:57:14 GMT"
},
{
"version": "v3",
"created": "Sat, 6 Jun 2015 16:58:41 GMT"
},
{
"version": "v4",
"created": "Fri, 12 Jun 2015 01:13:06 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Hu",
"Yuhuang",
""
],
[
"Ishwarya",
"M. S.",
""
],
[
"Loo",
"Chu Kiong",
""
]
] | TITLE: Classify Images with Conceptor Network
ABSTRACT: This article demonstrates a new conceptor network based classifier in
classifying images. Mathematical descriptions and analysis are presented.
Various tests are experimented using three benchmark datasets: MNIST, CIFAR-10
and CIFAR-100. The experiments displayed that conceptor network can offer
superior results and flexible configurations than conventional classifiers such
as Softmax Regression and Support Vector Machine (SVM).
|
1506.03837 | Weinan Zhang | Weinan Zhang, Jun Wang | Statistical Arbitrage Mining for Display Advertising | In the proceedings of the 21st ACM SIGKDD international conference on
Knowledge discovery and data mining (KDD 2015) | null | 10.1145/2783258.2783269 | null | cs.GT cs.DB | http://creativecommons.org/licenses/publicdomain/ | We study and formulate arbitrage in display advertising. Real-Time Bidding
(RTB) mimics stock spot exchanges and utilises computers to algorithmically buy
display ads per impression via a real-time auction. Despite the new automation,
the ad markets are still informationally inefficient due to the heavily
fragmented marketplaces. Two display impressions with similar or identical
effectiveness (e.g., measured by conversion or click-through rates for a
targeted audience) may sell for quite different prices at different market
segments or pricing schemes. In this paper, we propose a novel data mining
paradigm called Statistical Arbitrage Mining (SAM) focusing on mining and
exploiting price discrepancies between two pricing schemes. In essence, our
SAMer is a meta-bidder that hedges advertisers' risk between CPA (cost per
action)-based campaigns and CPM (cost per mille impressions)-based ad
inventories; it statistically assesses the potential profit and cost for an
incoming CPM bid request against a portfolio of CPA campaigns based on the
estimated conversion rate, bid landscape and other statistics learned from
historical data. In SAM, (i) functional optimisation is utilised to seek for
optimal bidding to maximise the expected arbitrage net profit, and (ii) a
portfolio-based risk management solution is leveraged to reallocate bid volume
and budget across the set of campaigns to make a risk and return trade-off. We
propose to jointly optimise both components in an EM fashion with high
efficiency to help the meta-bidder successfully catch the transient statistical
arbitrage opportunities in RTB. Both the offline experiments on a real-world
large-scale dataset and online A/B tests on a commercial platform demonstrate
the effectiveness of our proposed solution in exploiting arbitrage in various
model settings and market environments.
| [
{
"version": "v1",
"created": "Thu, 11 Jun 2015 21:05:26 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Zhang",
"Weinan",
""
],
[
"Wang",
"Jun",
""
]
] | TITLE: Statistical Arbitrage Mining for Display Advertising
ABSTRACT: We study and formulate arbitrage in display advertising. Real-Time Bidding
(RTB) mimics stock spot exchanges and utilises computers to algorithmically buy
display ads per impression via a real-time auction. Despite the new automation,
the ad markets are still informationally inefficient due to the heavily
fragmented marketplaces. Two display impressions with similar or identical
effectiveness (e.g., measured by conversion or click-through rates for a
targeted audience) may sell for quite different prices at different market
segments or pricing schemes. In this paper, we propose a novel data mining
paradigm called Statistical Arbitrage Mining (SAM) focusing on mining and
exploiting price discrepancies between two pricing schemes. In essence, our
SAMer is a meta-bidder that hedges advertisers' risk between CPA (cost per
action)-based campaigns and CPM (cost per mille impressions)-based ad
inventories; it statistically assesses the potential profit and cost for an
incoming CPM bid request against a portfolio of CPA campaigns based on the
estimated conversion rate, bid landscape and other statistics learned from
historical data. In SAM, (i) functional optimisation is utilised to seek for
optimal bidding to maximise the expected arbitrage net profit, and (ii) a
portfolio-based risk management solution is leveraged to reallocate bid volume
and budget across the set of campaigns to make a risk and return trade-off. We
propose to jointly optimise both components in an EM fashion with high
efficiency to help the meta-bidder successfully catch the transient statistical
arbitrage opportunities in RTB. Both the offline experiments on a real-world
large-scale dataset and online A/B tests on a commercial platform demonstrate
the effectiveness of our proposed solution in exploiting arbitrage in various
model settings and market environments.
|
1506.04051 | Lucia Maddalena | Lucia Maddalena and Alfredo Petrosino | Towards Benchmarking Scene Background Initialization | 6 pages, SBI dataset, SBMI2015 Workshop | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/3.0/ | Given a set of images of a scene taken at different times, the availability
of an initial background model that describes the scene without foreground
objects is the prerequisite for a wide range of applications, ranging from
video surveillance to computational photography. Even though several methods
have been proposed for scene background initialization, the lack of a common
groundtruthed dataset and of a common set of metrics makes it difficult to
compare their performance. To move first steps towards an easy and fair
comparison of these methods, we assembled a dataset of sequences frequently
adopted for background initialization, selected or created ground truths for
quantitative evaluation through a selected suite of metrics, and compared
results obtained by some existing methods, making all the material publicly
available.
| [
{
"version": "v1",
"created": "Fri, 12 Jun 2015 15:52:46 GMT"
}
] | 2015-06-15T00:00:00 | [
[
"Maddalena",
"Lucia",
""
],
[
"Petrosino",
"Alfredo",
""
]
] | TITLE: Towards Benchmarking Scene Background Initialization
ABSTRACT: Given a set of images of a scene taken at different times, the availability
of an initial background model that describes the scene without foreground
objects is the prerequisite for a wide range of applications, ranging from
video surveillance to computational photography. Even though several methods
have been proposed for scene background initialization, the lack of a common
groundtruthed dataset and of a common set of metrics makes it difficult to
compare their performance. To move first steps towards an easy and fair
comparison of these methods, we assembled a dataset of sequences frequently
adopted for background initialization, selected or created ground truths for
quantitative evaluation through a selected suite of metrics, and compared
results obtained by some existing methods, making all the material publicly
available.
|
1211.1073 | Venkat Chandrasekaran | Venkat Chandrasekaran and Michael I. Jordan | Computational and Statistical Tradeoffs via Convex Relaxation | null | null | 10.1073/pnas.1302293110 | null | math.ST cs.IT math.IT math.OC stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In modern data analysis, one is frequently faced with statistical inference
problems involving massive datasets. Processing such large datasets is usually
viewed as a substantial computational challenge. However, if data are a
statistician's main resource then access to more data should be viewed as an
asset rather than as a burden. In this paper we describe a computational
framework based on convex relaxation to reduce the computational complexity of
an inference procedure when one has access to increasingly larger datasets.
Convex relaxation techniques have been widely used in theoretical computer
science as they give tractable approximation algorithms to many computationally
intractable tasks. We demonstrate the efficacy of this methodology in
statistical estimation in providing concrete time-data tradeoffs in a class of
denoising problems. Thus, convex relaxation offers a principled approach to
exploit the statistical gains from larger datasets to reduce the runtime of
inference algorithms.
| [
{
"version": "v1",
"created": "Mon, 5 Nov 2012 23:28:44 GMT"
},
{
"version": "v2",
"created": "Mon, 26 Nov 2012 22:02:27 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Chandrasekaran",
"Venkat",
""
],
[
"Jordan",
"Michael I.",
""
]
] | TITLE: Computational and Statistical Tradeoffs via Convex Relaxation
ABSTRACT: In modern data analysis, one is frequently faced with statistical inference
problems involving massive datasets. Processing such large datasets is usually
viewed as a substantial computational challenge. However, if data are a
statistician's main resource then access to more data should be viewed as an
asset rather than as a burden. In this paper we describe a computational
framework based on convex relaxation to reduce the computational complexity of
an inference procedure when one has access to increasingly larger datasets.
Convex relaxation techniques have been widely used in theoretical computer
science as they give tractable approximation algorithms to many computationally
intractable tasks. We demonstrate the efficacy of this methodology in
statistical estimation in providing concrete time-data tradeoffs in a class of
denoising problems. Thus, convex relaxation offers a principled approach to
exploit the statistical gains from larger datasets to reduce the runtime of
inference algorithms.
|
1212.3333 | Ralf Kaehler | Ralf Kaehler, Tom Abel | Single-Pass GPU-Raycasting for Structured Adaptive Mesh Refinement Data | 12 pages, 7 figures. submitted to Visualization and Data Analysis
2013 | null | 10.1117/12.2008552 | null | astro-ph.IM cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Structured Adaptive Mesh Refinement (SAMR) is a popular numerical technique
to study processes with high spatial and temporal dynamic range. It reduces
computational requirements by adapting the lattice on which the underlying
differential equations are solved to most efficiently represent the solution.
Particularly in astrophysics and cosmology such simulations now can capture
spatial scales ten orders of magnitude apart and more. The irregular locations
and extensions of the refined regions in the SAMR scheme and the fact that
different resolution levels partially overlap, poses a challenge for GPU-based
direct volume rendering methods. kD-trees have proven to be advantageous to
subdivide the data domain into non-overlapping blocks of equally sized cells,
optimal for the texture units of current graphics hardware, but previous
GPU-supported raycasting approaches for SAMR data using this data structure
required a separate rendering pass for each node, preventing the application of
many advanced lighting schemes that require simultaneous access to more than
one block of cells. In this paper we present a single-pass GPU-raycasting
algorithm for SAMR data that is based on a kD-tree. The tree is efficiently
encoded by a set of 3D-textures, which allows to adaptively sample complete
rays entirely on the GPU without any CPU interaction. We discuss two different
data storage strategies to access the grid data on the GPU and apply them to
several datasets to prove the benefits of the proposed method.
| [
{
"version": "v1",
"created": "Thu, 13 Dec 2012 21:00:02 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Kaehler",
"Ralf",
""
],
[
"Abel",
"Tom",
""
]
] | TITLE: Single-Pass GPU-Raycasting for Structured Adaptive Mesh Refinement Data
ABSTRACT: Structured Adaptive Mesh Refinement (SAMR) is a popular numerical technique
to study processes with high spatial and temporal dynamic range. It reduces
computational requirements by adapting the lattice on which the underlying
differential equations are solved to most efficiently represent the solution.
Particularly in astrophysics and cosmology such simulations now can capture
spatial scales ten orders of magnitude apart and more. The irregular locations
and extensions of the refined regions in the SAMR scheme and the fact that
different resolution levels partially overlap, poses a challenge for GPU-based
direct volume rendering methods. kD-trees have proven to be advantageous to
subdivide the data domain into non-overlapping blocks of equally sized cells,
optimal for the texture units of current graphics hardware, but previous
GPU-supported raycasting approaches for SAMR data using this data structure
required a separate rendering pass for each node, preventing the application of
many advanced lighting schemes that require simultaneous access to more than
one block of cells. In this paper we present a single-pass GPU-raycasting
algorithm for SAMR data that is based on a kD-tree. The tree is efficiently
encoded by a set of 3D-textures, which allows to adaptively sample complete
rays entirely on the GPU without any CPU interaction. We discuss two different
data storage strategies to access the grid data on the GPU and apply them to
several datasets to prove the benefits of the proposed method.
|
1301.4546 | Akira Kageyama | Akira Kageyama and Tomoki Yamada | An Approach to Exascale Visualization: Interactive Viewing of In-Situ
Visualization | Will appear in Comput. Phys. Comm | null | 10.1016/j.cpc.2013.08.017 | null | physics.comp-ph cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the coming era of exascale supercomputing, in-situ visualization will be a
crucial approach for reducing the output data size. A problem of in-situ
visualization is that it loses interactivity if a steering method is not
adopted. In this paper, we propose a new method for the interactive analysis of
in-situ visualization images produced by a batch simulation job. A key idea is
to apply numerous (thousands to millions) in-situ visualizations
simultaneously. The viewer then analyzes the image database interactively
during postprocessing. If each movie can be compressed to 100 MB, one million
movies will only require 100 TB, which is smaller than the size of the raw
numerical data in exascale supercomputing. We performed a feasibility study
using the proposed method. Multiple movie files were produced by a simulation
and they were analyzed using a specially designed movie player. The user could
change the viewing angle, the visualization method, and the parameters
interactively by retrieving an appropriate sequence of images from the movie
dataset.
| [
{
"version": "v1",
"created": "Sat, 19 Jan 2013 08:39:58 GMT"
},
{
"version": "v2",
"created": "Sat, 24 Aug 2013 00:21:14 GMT"
},
{
"version": "v3",
"created": "Fri, 13 Sep 2013 08:55:28 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Kageyama",
"Akira",
""
],
[
"Yamada",
"Tomoki",
""
]
] | TITLE: An Approach to Exascale Visualization: Interactive Viewing of In-Situ
Visualization
ABSTRACT: In the coming era of exascale supercomputing, in-situ visualization will be a
crucial approach for reducing the output data size. A problem of in-situ
visualization is that it loses interactivity if a steering method is not
adopted. In this paper, we propose a new method for the interactive analysis of
in-situ visualization images produced by a batch simulation job. A key idea is
to apply numerous (thousands to millions) in-situ visualizations
simultaneously. The viewer then analyzes the image database interactively
during postprocessing. If each movie can be compressed to 100 MB, one million
movies will only require 100 TB, which is smaller than the size of the raw
numerical data in exascale supercomputing. We performed a feasibility study
using the proposed method. Multiple movie files were produced by a simulation
and they were analyzed using a specially designed movie player. The user could
change the viewing angle, the visualization method, and the parameters
interactively by retrieving an appropriate sequence of images from the movie
dataset.
|
1301.6600 | Tao Wang Dr. | Tao Wang, Francois Glineur, Jerome Louveaux, Luc Vandendorpe | Weighted Sum Rate Maximization for Downlink OFDMA with Subcarrier-pair
based Opportunistic DF Relaying | 8 figures, accepted and to be published in IEEE Transactions on
Signal Processing. arXiv admin note: text overlap with arXiv:1301.2935 | null | 10.1109/TSP.2013.2245326 | null | cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses a weighted sum rate (WSR) maximization problem for
downlink OFDMA aided by a decode-and-forward (DF) relay under a total power
constraint. A novel subcarrier-pair based opportunistic DF relaying protocol is
proposed. Specifically, user message bits are transmitted in two time slots. A
subcarrier in the first slot can be paired with a subcarrier in the second slot
for the DF relay-aided transmission to a user. In particular, the source and
the relay can transmit simultaneously to implement beamforming at the
subcarrier in the second slot. Each unpaired subcarrier in either the first or
second slot is used for the source's direct transmission to a user. A benchmark
protocol, same as the proposed one except that the transmit beamforming is not
used for the relay-aided transmission, is also considered. For each protocol, a
polynomial-complexity algorithm is developed to find at least an approximately
optimum resource allocation (RA), by using continuous relaxation, the dual
method, and Hungarian algorithm. Instrumental to the algorithm design is an
elegant definition of optimization variables, motivated by the idea of
regarding the unpaired subcarriers as virtual subcarrier pairs in the direct
transmission mode. The effectiveness of the RA algorithm and the impact of
relay position and total power on the protocols' performance are illustrated by
numerical experiments. The proposed protocol always leads to a maximum WSR
equal to or greater than that for the benchmark one, and the performance gain
of using the proposed one is significant especially when the relay is in close
proximity to the source and the total power is low. Theoretical analysis is
presented to interpret these observations.
| [
{
"version": "v1",
"created": "Mon, 28 Jan 2013 16:54:21 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Wang",
"Tao",
""
],
[
"Glineur",
"Francois",
""
],
[
"Louveaux",
"Jerome",
""
],
[
"Vandendorpe",
"Luc",
""
]
] | TITLE: Weighted Sum Rate Maximization for Downlink OFDMA with Subcarrier-pair
based Opportunistic DF Relaying
ABSTRACT: This paper addresses a weighted sum rate (WSR) maximization problem for
downlink OFDMA aided by a decode-and-forward (DF) relay under a total power
constraint. A novel subcarrier-pair based opportunistic DF relaying protocol is
proposed. Specifically, user message bits are transmitted in two time slots. A
subcarrier in the first slot can be paired with a subcarrier in the second slot
for the DF relay-aided transmission to a user. In particular, the source and
the relay can transmit simultaneously to implement beamforming at the
subcarrier in the second slot. Each unpaired subcarrier in either the first or
second slot is used for the source's direct transmission to a user. A benchmark
protocol, same as the proposed one except that the transmit beamforming is not
used for the relay-aided transmission, is also considered. For each protocol, a
polynomial-complexity algorithm is developed to find at least an approximately
optimum resource allocation (RA), by using continuous relaxation, the dual
method, and Hungarian algorithm. Instrumental to the algorithm design is an
elegant definition of optimization variables, motivated by the idea of
regarding the unpaired subcarriers as virtual subcarrier pairs in the direct
transmission mode. The effectiveness of the RA algorithm and the impact of
relay position and total power on the protocols' performance are illustrated by
numerical experiments. The proposed protocol always leads to a maximum WSR
equal to or greater than that for the benchmark one, and the performance gain
of using the proposed one is significant especially when the relay is in close
proximity to the source and the total power is low. Theoretical analysis is
presented to interpret these observations.
|
1409.5400 | Tobias Weyand | Tobias Weyand and Bastian Leibe | Visual Landmark Recognition from Internet Photo Collections: A
Large-Scale Evaluation | null | null | 10.1016/j.cviu.2015.02.002 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The task of a visual landmark recognition system is to identify photographed
buildings or objects in query photos and to provide the user with relevant
information on them. With their increasing coverage of the world's landmark
buildings and objects, Internet photo collections are now being used as a
source for building such systems in a fully automatic fashion. This process
typically consists of three steps: clustering large amounts of images by the
objects they depict; determining object names from user-provided tags; and
building a robust, compact, and efficient recognition index. To this date,
however, there is little empirical information on how well current approaches
for those steps perform in a large-scale open-set mining and recognition task.
Furthermore, there is little empirical information on how recognition
performance varies for different types of landmark objects and where there is
still potential for improvement. With this paper, we intend to fill these gaps.
Using a dataset of 500k images from Paris, we analyze each component of the
landmark recognition pipeline in order to answer the following questions: How
many and what kinds of objects can be discovered automatically? How can we best
use the resulting image clusters to recognize the object in a query? How can
the object be efficiently represented in memory for recognition? How reliably
can semantic information be extracted? And finally: What are the limiting
factors in the resulting pipeline from query to semantics? We evaluate how
different choices of methods and parameters for the individual pipeline steps
affect overall system performance and examine their effects for different query
categories such as buildings, paintings or sculptures.
| [
{
"version": "v1",
"created": "Thu, 18 Sep 2014 18:28:20 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Weyand",
"Tobias",
""
],
[
"Leibe",
"Bastian",
""
]
] | TITLE: Visual Landmark Recognition from Internet Photo Collections: A
Large-Scale Evaluation
ABSTRACT: The task of a visual landmark recognition system is to identify photographed
buildings or objects in query photos and to provide the user with relevant
information on them. With their increasing coverage of the world's landmark
buildings and objects, Internet photo collections are now being used as a
source for building such systems in a fully automatic fashion. This process
typically consists of three steps: clustering large amounts of images by the
objects they depict; determining object names from user-provided tags; and
building a robust, compact, and efficient recognition index. To this date,
however, there is little empirical information on how well current approaches
for those steps perform in a large-scale open-set mining and recognition task.
Furthermore, there is little empirical information on how recognition
performance varies for different types of landmark objects and where there is
still potential for improvement. With this paper, we intend to fill these gaps.
Using a dataset of 500k images from Paris, we analyze each component of the
landmark recognition pipeline in order to answer the following questions: How
many and what kinds of objects can be discovered automatically? How can we best
use the resulting image clusters to recognize the object in a query? How can
the object be efficiently represented in memory for recognition? How reliably
can semantic information be extracted? And finally: What are the limiting
factors in the resulting pipeline from query to semantics? We evaluate how
different choices of methods and parameters for the individual pipeline steps
affect overall system performance and examine their effects for different query
categories such as buildings, paintings or sculptures.
|
1412.6632 | Junhua Mao | Junhua Mao, Wei Xu, Yi Yang, Jiang Wang, Zhiheng Huang, Alan Yuille | Deep Captioning with Multimodal Recurrent Neural Networks (m-RNN) | Add a simple strategy to boost the performance of image captioning
task significantly. More details are shown in Section 8 of the paper. The
code and related data are available at https://github.com/mjhucla/mRNN-CR ;.
arXiv admin note: substantial text overlap with arXiv:1410.1090 | ICLR 2015 | null | null | cs.CV cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present a multimodal Recurrent Neural Network (m-RNN) model
for generating novel image captions. It directly models the probability
distribution of generating a word given previous words and an image. Image
captions are generated by sampling from this distribution. The model consists
of two sub-networks: a deep recurrent neural network for sentences and a deep
convolutional network for images. These two sub-networks interact with each
other in a multimodal layer to form the whole m-RNN model. The effectiveness of
our model is validated on four benchmark datasets (IAPR TC-12, Flickr 8K,
Flickr 30K and MS COCO). Our model outperforms the state-of-the-art methods. In
addition, we apply the m-RNN model to retrieval tasks for retrieving images or
sentences, and achieves significant performance improvement over the
state-of-the-art methods which directly optimize the ranking objective function
for retrieval. The project page of this work is:
www.stat.ucla.edu/~junhua.mao/m-RNN.html .
| [
{
"version": "v1",
"created": "Sat, 20 Dec 2014 08:10:04 GMT"
},
{
"version": "v2",
"created": "Fri, 26 Dec 2014 08:24:11 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Mar 2015 04:17:48 GMT"
},
{
"version": "v4",
"created": "Fri, 10 Apr 2015 21:03:35 GMT"
},
{
"version": "v5",
"created": "Thu, 11 Jun 2015 15:26:58 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Mao",
"Junhua",
""
],
[
"Xu",
"Wei",
""
],
[
"Yang",
"Yi",
""
],
[
"Wang",
"Jiang",
""
],
[
"Huang",
"Zhiheng",
""
],
[
"Yuille",
"Alan",
""
]
] | TITLE: Deep Captioning with Multimodal Recurrent Neural Networks (m-RNN)
ABSTRACT: In this paper, we present a multimodal Recurrent Neural Network (m-RNN) model
for generating novel image captions. It directly models the probability
distribution of generating a word given previous words and an image. Image
captions are generated by sampling from this distribution. The model consists
of two sub-networks: a deep recurrent neural network for sentences and a deep
convolutional network for images. These two sub-networks interact with each
other in a multimodal layer to form the whole m-RNN model. The effectiveness of
our model is validated on four benchmark datasets (IAPR TC-12, Flickr 8K,
Flickr 30K and MS COCO). Our model outperforms the state-of-the-art methods. In
addition, we apply the m-RNN model to retrieval tasks for retrieving images or
sentences, and achieves significant performance improvement over the
state-of-the-art methods which directly optimize the ranking objective function
for retrieval. The project page of this work is:
www.stat.ucla.edu/~junhua.mao/m-RNN.html .
|
1502.02921 | Albert Saa-Garriga | Albert Saa-Garriga, David Castells-Rufas and Jordi Carrabina | OMP2MPI: Automatic MPI code generation from OpenMP programs | Presented at HIP3ES, 2015 (arXiv: 1501.03064) | null | null | HIP3ES/2015/06 | cs.DC cs.PF cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present OMP2MPI a tool that generates automatically MPI
source code from OpenMP. With this transformation the original program can be
adapted to be able to exploit a larger number of processors by surpassing the
limits of the node level on large HPC clusters. The transformation can also be
useful to adapt the source code to execute in distributed memory many-cores
with message passing support. In addition, the resulting MPI code can be used
as an starting point that still can be further optimized by software engineers.
The transformation process is focused on detecting OpenMP parallel loops and
distributing them in a master/worker pattern. A set of micro-benchmarks have
been used to verify the correctness of the the transformation and to measure
the resulting performance. Surprisingly not only the automatically generated
code is correct by construction, but also it often performs faster even when
executed with MPI.
| [
{
"version": "v1",
"created": "Tue, 10 Feb 2015 14:32:25 GMT"
},
{
"version": "v2",
"created": "Thu, 11 Jun 2015 08:52:41 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Saa-Garriga",
"Albert",
""
],
[
"Castells-Rufas",
"David",
""
],
[
"Carrabina",
"Jordi",
""
]
] | TITLE: OMP2MPI: Automatic MPI code generation from OpenMP programs
ABSTRACT: In this paper, we present OMP2MPI a tool that generates automatically MPI
source code from OpenMP. With this transformation the original program can be
adapted to be able to exploit a larger number of processors by surpassing the
limits of the node level on large HPC clusters. The transformation can also be
useful to adapt the source code to execute in distributed memory many-cores
with message passing support. In addition, the resulting MPI code can be used
as an starting point that still can be further optimized by software engineers.
The transformation process is focused on detecting OpenMP parallel loops and
distributing them in a master/worker pattern. A set of micro-benchmarks have
been used to verify the correctness of the the transformation and to measure
the resulting performance. Surprisingly not only the automatically generated
code is correct by construction, but also it often performs faster even when
executed with MPI.
|
1506.03623 | Han Xiao Bookman | Han Xiao, Xiaoyan Zhu | Max-Entropy Feed-Forward Clustering Neural Network | This paper has been published in ICANN 2015 | ICANN 2015: International Conference on Artificial Neural
Networks, Amsterdam, The Netherlands, (May 14-15, 2015) | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The outputs of non-linear feed-forward neural network are positive, which
could be treated as probability when they are normalized to one. If we take
Entropy-Based Principle into consideration, the outputs for each sample could
be represented as the distribution of this sample for different clusters.
Entropy-Based Principle is the principle with which we could estimate the
unknown distribution under some limited conditions. As this paper defines two
processes in Feed-Forward Neural Network, our limited condition is the
abstracted features of samples which are worked out in the abstraction process.
And the final outputs are the probability distribution for different clusters
in the clustering process. As Entropy-Based Principle is considered into the
feed-forward neural network, a clustering method is born. We have conducted
some experiments on six open UCI datasets, comparing with a few baselines and
applied purity as the measurement . The results illustrate that our method
outperforms all the other baselines that are most popular clustering methods.
| [
{
"version": "v1",
"created": "Thu, 11 Jun 2015 11:01:40 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Xiao",
"Han",
""
],
[
"Zhu",
"Xiaoyan",
""
]
] | TITLE: Max-Entropy Feed-Forward Clustering Neural Network
ABSTRACT: The outputs of non-linear feed-forward neural network are positive, which
could be treated as probability when they are normalized to one. If we take
Entropy-Based Principle into consideration, the outputs for each sample could
be represented as the distribution of this sample for different clusters.
Entropy-Based Principle is the principle with which we could estimate the
unknown distribution under some limited conditions. As this paper defines two
processes in Feed-Forward Neural Network, our limited condition is the
abstracted features of samples which are worked out in the abstraction process.
And the final outputs are the probability distribution for different clusters
in the clustering process. As Entropy-Based Principle is considered into the
feed-forward neural network, a clustering method is born. We have conducted
some experiments on six open UCI datasets, comparing with a few baselines and
applied purity as the measurement . The results illustrate that our method
outperforms all the other baselines that are most popular clustering methods.
|
1506.03626 | Han Xiao Bookman | Han Xiao, Xiaoyan Zhu | Margin-Based Feed-Forward Neural Network Classifiers | This paper has been published in ICANN 2015: International Conference
on Artificial Neural Networks, Amsterdam, The Netherlands, (May 14-15, 2015) | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Margin-Based Principle has been proposed for a long time, it has been proved
that this principle could reduce the structural risk and improve the
performance in both theoretical and practical aspects. Meanwhile, feed-forward
neural network is a traditional classifier, which is very hot at present with a
deeper architecture. However, the training algorithm of feed-forward neural
network is developed and generated from Widrow-Hoff Principle that means to
minimize the squared error. In this paper, we propose a new training algorithm
for feed-forward neural networks based on Margin-Based Principle, which could
effectively promote the accuracy and generalization ability of neural network
classifiers with less labelled samples and flexible network. We have conducted
experiments on four UCI open datasets and achieved good results as expected. In
conclusion, our model could handle more sparse labelled and more high-dimension
dataset in a high accuracy while modification from old ANN method to our method
is easy and almost free of work.
| [
{
"version": "v1",
"created": "Thu, 11 Jun 2015 11:10:25 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Xiao",
"Han",
""
],
[
"Zhu",
"Xiaoyan",
""
]
] | TITLE: Margin-Based Feed-Forward Neural Network Classifiers
ABSTRACT: Margin-Based Principle has been proposed for a long time, it has been proved
that this principle could reduce the structural risk and improve the
performance in both theoretical and practical aspects. Meanwhile, feed-forward
neural network is a traditional classifier, which is very hot at present with a
deeper architecture. However, the training algorithm of feed-forward neural
network is developed and generated from Widrow-Hoff Principle that means to
minimize the squared error. In this paper, we propose a new training algorithm
for feed-forward neural networks based on Margin-Based Principle, which could
effectively promote the accuracy and generalization ability of neural network
classifiers with less labelled samples and flexible network. We have conducted
experiments on four UCI open datasets and achieved good results as expected. In
conclusion, our model could handle more sparse labelled and more high-dimension
dataset in a high accuracy while modification from old ANN method to our method
is easy and almost free of work.
|
1506.03668 | Zolzaya Dashdorj | Zolzaya Dashdorj and Stanislav Sobolevsky | Impact of the spatial context on human communication activity | 12 pages, 11 figures | null | null | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Technology development produces terabytes of data generated by hu- man
activity in space and time. This enormous amount of data often called big data
becomes crucial for delivering new insights to decision makers. It contains
behavioral information on different types of human activity influenced by many
external factors such as geographic infor- mation and weather forecast. Early
recognition and prediction of those human behaviors are of great importance in
many societal applications like health-care, risk management and urban
planning, etc. In this pa- per, we investigate relevant geographical areas
based on their categories of human activities (i.e., working and shopping)
which identified from ge- ographic information (i.e., Openstreetmap). We use
spectral clustering followed by k-means clustering algorithm based on TF/IDF
cosine simi- larity metric. We evaluate the quality of those observed clusters
with the use of silhouette coefficients which are estimated based on the
similari- ties of the mobile communication activity temporal patterns. The area
clusters are further used to explain typical or exceptional communication
activities. We demonstrate the study using a real dataset containing 1 million
Call Detailed Records. This type of analysis and its application are important
for analyzing the dependency of human behaviors from the external factors and
hidden relationships and unknown correlations and other useful information that
can support decision-making.
| [
{
"version": "v1",
"created": "Thu, 11 Jun 2015 13:46:16 GMT"
}
] | 2015-06-12T00:00:00 | [
[
"Dashdorj",
"Zolzaya",
""
],
[
"Sobolevsky",
"Stanislav",
""
]
] | TITLE: Impact of the spatial context on human communication activity
ABSTRACT: Technology development produces terabytes of data generated by hu- man
activity in space and time. This enormous amount of data often called big data
becomes crucial for delivering new insights to decision makers. It contains
behavioral information on different types of human activity influenced by many
external factors such as geographic infor- mation and weather forecast. Early
recognition and prediction of those human behaviors are of great importance in
many societal applications like health-care, risk management and urban
planning, etc. In this pa- per, we investigate relevant geographical areas
based on their categories of human activities (i.e., working and shopping)
which identified from ge- ographic information (i.e., Openstreetmap). We use
spectral clustering followed by k-means clustering algorithm based on TF/IDF
cosine simi- larity metric. We evaluate the quality of those observed clusters
with the use of silhouette coefficients which are estimated based on the
similari- ties of the mobile communication activity temporal patterns. The area
clusters are further used to explain typical or exceptional communication
activities. We demonstrate the study using a real dataset containing 1 million
Call Detailed Records. This type of analysis and its application are important
for analyzing the dependency of human behaviors from the external factors and
hidden relationships and unknown correlations and other useful information that
can support decision-making.
|
1210.1095 | Francesco Vezzi | Francesco Vezzi, Giuseppe Narzisi and Bud Mishra | Reevaluating Assembly Evaluations with Feature Response Curves: GAGE and
Assemblathons | Submitted to PLoS One. Supplementary material available at
http://www.nada.kth.se/~vezzi/publications/supplementary.pdf and
http://cs.nyu.edu/mishra/PUBLICATIONS/12.supplementaryFRC.pdf | null | 10.1371/journal.pone.0052210 | null | q-bio.GN cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In just the last decade, a multitude of bio-technologies and software
pipelines have emerged to revolutionize genomics. To further their central
goal, they aim to accelerate and improve the quality of de novo whole-genome
assembly starting from short DNA reads. However, the performance of each of
these tools is contingent on the length and quality of the sequencing data, the
structure and complexity of the genome sequence, and the resolution and quality
of long-range information. Furthermore, in the absence of any metric that
captures the most fundamental "features" of a high-quality assembly, there is
no obvious recipe for users to select the most desirable assembler/assembly.
International competitions such as Assemblathons or GAGE tried to identify the
best assembler(s) and their features. Some what circuitously, the only
available approach to gauge de novo assemblies and assemblers relies solely on
the availability of a high-quality fully assembled reference genome sequence.
Still worse, reference-guided evaluations are often both difficult to analyze,
leading to conclusions that are difficult to interpret. In this paper, we
circumvent many of these issues by relying upon a tool, dubbed FRCbam, which is
capable of evaluating de novo assemblies from the read-layouts even when no
reference exists. We extend the FRCurve approach to cases where lay-out
information may have been obscured, as is true in many deBruijn-graph-based
algorithms. As a by-product, FRCurve now expands its applicability to a much
wider class of assemblers -- thus, identifying higher-quality members of this
group, their inter-relations as well as sensitivity to carefully selected
features, with or without the support of a reference sequence or layout for the
reads. The paper concludes by reevaluating several recently conducted assembly
competitions and the datasets that have resulted from them.
| [
{
"version": "v1",
"created": "Wed, 3 Oct 2012 13:02:30 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Vezzi",
"Francesco",
""
],
[
"Narzisi",
"Giuseppe",
""
],
[
"Mishra",
"Bud",
""
]
] | TITLE: Reevaluating Assembly Evaluations with Feature Response Curves: GAGE and
Assemblathons
ABSTRACT: In just the last decade, a multitude of bio-technologies and software
pipelines have emerged to revolutionize genomics. To further their central
goal, they aim to accelerate and improve the quality of de novo whole-genome
assembly starting from short DNA reads. However, the performance of each of
these tools is contingent on the length and quality of the sequencing data, the
structure and complexity of the genome sequence, and the resolution and quality
of long-range information. Furthermore, in the absence of any metric that
captures the most fundamental "features" of a high-quality assembly, there is
no obvious recipe for users to select the most desirable assembler/assembly.
International competitions such as Assemblathons or GAGE tried to identify the
best assembler(s) and their features. Some what circuitously, the only
available approach to gauge de novo assemblies and assemblers relies solely on
the availability of a high-quality fully assembled reference genome sequence.
Still worse, reference-guided evaluations are often both difficult to analyze,
leading to conclusions that are difficult to interpret. In this paper, we
circumvent many of these issues by relying upon a tool, dubbed FRCbam, which is
capable of evaluating de novo assemblies from the read-layouts even when no
reference exists. We extend the FRCurve approach to cases where lay-out
information may have been obscured, as is true in many deBruijn-graph-based
algorithms. As a by-product, FRCurve now expands its applicability to a much
wider class of assemblers -- thus, identifying higher-quality members of this
group, their inter-relations as well as sensitivity to carefully selected
features, with or without the support of a reference sequence or layout for the
reads. The paper concludes by reevaluating several recently conducted assembly
competitions and the datasets that have resulted from them.
|
1404.0466 | Da Kuang | Da Kuang, Alex Gittens, Raffay Hamid | piCholesky: Polynomial Interpolation of Multiple Cholesky Factors for
Efficient Approximate Cross-Validation | null | null | null | null | cs.LG cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The dominant cost in solving least-square problems using Newton's method is
often that of factorizing the Hessian matrix over multiple values of the
regularization parameter ($\lambda$). We propose an efficient way to
interpolate the Cholesky factors of the Hessian matrix computed over a small
set of $\lambda$ values. This approximation enables us to optimally minimize
the hold-out error while incurring only a fraction of the cost compared to
exact cross-validation. We provide a formal error bound for our approximation
scheme and present solutions to a set of key implementation challenges that
allow our approach to maximally exploit the compute power of modern
architectures. We present a thorough empirical analysis over multiple datasets
to show the effectiveness of our approach.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2014 05:33:41 GMT"
},
{
"version": "v2",
"created": "Wed, 10 Jun 2015 18:20:16 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Kuang",
"Da",
""
],
[
"Gittens",
"Alex",
""
],
[
"Hamid",
"Raffay",
""
]
] | TITLE: piCholesky: Polynomial Interpolation of Multiple Cholesky Factors for
Efficient Approximate Cross-Validation
ABSTRACT: The dominant cost in solving least-square problems using Newton's method is
often that of factorizing the Hessian matrix over multiple values of the
regularization parameter ($\lambda$). We propose an efficient way to
interpolate the Cholesky factors of the Hessian matrix computed over a small
set of $\lambda$ values. This approximation enables us to optimally minimize
the hold-out error while incurring only a fraction of the cost compared to
exact cross-validation. We provide a formal error bound for our approximation
scheme and present solutions to a set of key implementation challenges that
allow our approach to maximally exploit the compute power of modern
architectures. We present a thorough empirical analysis over multiple datasets
to show the effectiveness of our approach.
|
1412.7449 | Oriol Vinyals | Oriol Vinyals, Lukasz Kaiser, Terry Koo, Slav Petrov, Ilya Sutskever,
Geoffrey Hinton | Grammar as a Foreign Language | null | null | null | null | cs.CL cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Syntactic constituency parsing is a fundamental problem in natural language
processing and has been the subject of intensive research and engineering for
decades. As a result, the most accurate parsers are domain specific, complex,
and inefficient. In this paper we show that the domain agnostic
attention-enhanced sequence-to-sequence model achieves state-of-the-art results
on the most widely used syntactic constituency parsing dataset, when trained on
a large synthetic corpus that was annotated using existing parsers. It also
matches the performance of standard parsers when trained only on a small
human-annotated dataset, which shows that this model is highly data-efficient,
in contrast to sequence-to-sequence models without the attention mechanism. Our
parser is also fast, processing over a hundred sentences per second with an
unoptimized CPU implementation.
| [
{
"version": "v1",
"created": "Tue, 23 Dec 2014 17:16:24 GMT"
},
{
"version": "v2",
"created": "Sat, 28 Feb 2015 03:16:54 GMT"
},
{
"version": "v3",
"created": "Tue, 9 Jun 2015 22:41:07 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Vinyals",
"Oriol",
""
],
[
"Kaiser",
"Lukasz",
""
],
[
"Koo",
"Terry",
""
],
[
"Petrov",
"Slav",
""
],
[
"Sutskever",
"Ilya",
""
],
[
"Hinton",
"Geoffrey",
""
]
] | TITLE: Grammar as a Foreign Language
ABSTRACT: Syntactic constituency parsing is a fundamental problem in natural language
processing and has been the subject of intensive research and engineering for
decades. As a result, the most accurate parsers are domain specific, complex,
and inefficient. In this paper we show that the domain agnostic
attention-enhanced sequence-to-sequence model achieves state-of-the-art results
on the most widely used syntactic constituency parsing dataset, when trained on
a large synthetic corpus that was annotated using existing parsers. It also
matches the performance of standard parsers when trained only on a small
human-annotated dataset, which shows that this model is highly data-efficient,
in contrast to sequence-to-sequence models without the attention mechanism. Our
parser is also fast, processing over a hundred sentences per second with an
unoptimized CPU implementation.
|
1506.01973 | Hyungyu Shin | Jinha Kim (1 and 2), Hyungyu Shin (1), Wook-Shin Han (1), Sungpack
Hong (2), and Hassan Chafi (2) ((1) POSTECH, South Korea, (2) Oracle Labs,
USA) | Taming Subgraph Isomorphism for RDF Query Processing | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | RDF data are used to model knowledge in various areas such as life sciences,
Semantic Web, bioinformatics, and social graphs. The size of real RDF data
reaches billions of triples. This calls for a framework for efficiently
processing RDF data. The core function of processing RDF data is subgraph
pattern matching. There have been two completely different directions for
supporting efficient subgraph pattern matching. One direction is to develop
specialized RDF query processing engines exploiting the properties of RDF data
for the last decade, while the other direction is to develop efficient subgraph
isomorphism algorithms for general, labeled graphs for over 30 years. Although
both directions have a similar goal (i.e., finding subgraphs in data graphs for
a given query graph), they have been independently researched without clear
reason. We argue that a subgraph isomorphism algorithm can be easily modified
to handle the graph homomorphism, which is the RDF pattern matching semantics,
by just removing the injectivity constraint. In this paper, based on the
state-of-the-art subgraph isomorphism algorithm, we propose an in-memory
solution, TurboHOM++, which is tamed for the RDF processing, and we compare it
with the representative RDF processing engines for several RDF benchmarks in a
server machine where billions of triples can be loaded in memory. In order to
speed up TurboHOM++, we also provide a simple yet effective transformation and
a series of optimization techniques. Extensive experiments using several RDF
benchmarks show that TurboHOM++ consistently and significantly outperforms the
representative RDF engines. Specifically, TurboHOM++ outperforms its
competitors by up to five orders of magnitude.
| [
{
"version": "v1",
"created": "Fri, 5 Jun 2015 17:00:47 GMT"
},
{
"version": "v2",
"created": "Wed, 10 Jun 2015 08:54:48 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Kim",
"Jinha",
"",
"1 and 2"
],
[
"Shin",
"Hyungyu",
""
],
[
"Han",
"Wook-Shin",
""
],
[
"Hong",
"Sungpack",
""
],
[
"Chafi",
"Hassan",
""
]
] | TITLE: Taming Subgraph Isomorphism for RDF Query Processing
ABSTRACT: RDF data are used to model knowledge in various areas such as life sciences,
Semantic Web, bioinformatics, and social graphs. The size of real RDF data
reaches billions of triples. This calls for a framework for efficiently
processing RDF data. The core function of processing RDF data is subgraph
pattern matching. There have been two completely different directions for
supporting efficient subgraph pattern matching. One direction is to develop
specialized RDF query processing engines exploiting the properties of RDF data
for the last decade, while the other direction is to develop efficient subgraph
isomorphism algorithms for general, labeled graphs for over 30 years. Although
both directions have a similar goal (i.e., finding subgraphs in data graphs for
a given query graph), they have been independently researched without clear
reason. We argue that a subgraph isomorphism algorithm can be easily modified
to handle the graph homomorphism, which is the RDF pattern matching semantics,
by just removing the injectivity constraint. In this paper, based on the
state-of-the-art subgraph isomorphism algorithm, we propose an in-memory
solution, TurboHOM++, which is tamed for the RDF processing, and we compare it
with the representative RDF processing engines for several RDF benchmarks in a
server machine where billions of triples can be loaded in memory. In order to
speed up TurboHOM++, we also provide a simple yet effective transformation and
a series of optimization techniques. Extensive experiments using several RDF
benchmarks show that TurboHOM++ consistently and significantly outperforms the
representative RDF engines. Specifically, TurboHOM++ outperforms its
competitors by up to five orders of magnitude.
|
1506.03139 | Keenon Werling | Keenon Werling, Gabor Angeli, Christopher Manning | Robust Subgraph Generation Improves Abstract Meaning Representation
Parsing | To appear in ACL 2015 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Abstract Meaning Representation (AMR) is a representation for open-domain
rich semantics, with potential use in fields like event extraction and machine
translation. Node generation, typically done using a simple dictionary lookup,
is currently an important limiting factor in AMR parsing. We propose a small
set of actions that derive AMR subgraphs by transformations on spans of text,
which allows for more robust learning of this stage. Our set of construction
actions generalize better than the previous approach, and can be learned with a
simple classifier. We improve on the previous state-of-the-art result for AMR
parsing, boosting end-to-end performance by 3 F$_1$ on both the LDC2013E117 and
LDC2014T12 datasets.
| [
{
"version": "v1",
"created": "Wed, 10 Jun 2015 00:40:12 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Werling",
"Keenon",
""
],
[
"Angeli",
"Gabor",
""
],
[
"Manning",
"Christopher",
""
]
] | TITLE: Robust Subgraph Generation Improves Abstract Meaning Representation
Parsing
ABSTRACT: The Abstract Meaning Representation (AMR) is a representation for open-domain
rich semantics, with potential use in fields like event extraction and machine
translation. Node generation, typically done using a simple dictionary lookup,
is currently an important limiting factor in AMR parsing. We propose a small
set of actions that derive AMR subgraphs by transformations on spans of text,
which allows for more robust learning of this stage. Our set of construction
actions generalize better than the previous approach, and can be learned with a
simple classifier. We improve on the previous state-of-the-art result for AMR
parsing, boosting end-to-end performance by 3 F$_1$ on both the LDC2013E117 and
LDC2014T12 datasets.
|
1506.03184 | Cong Yao | Xinyu Zhou and Shuchang Zhou and Cong Yao and Zhimin Cao and Qi Yin | ICDAR 2015 Text Reading in the Wild Competition | 3 pages, 2 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, text detection and recognition in natural scenes are becoming
increasing popular in the computer vision community as well as the document
analysis community. However, majority of the existing ideas, algorithms and
systems are specifically designed for English. This technical report presents
the final results of the ICDAR 2015 Text Reading in the Wild (TRW 2015)
competition, which aims at establishing a benchmark for assessing detection and
recognition algorithms devised for both Chinese and English scripts and
providing a playground for researchers from the community. In this article, we
describe in detail the dataset, tasks, evaluation protocols and participants of
this competition, and report the performance of the participating methods.
Moreover, promising directions for future research are discussed.
| [
{
"version": "v1",
"created": "Wed, 10 Jun 2015 06:46:55 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Zhou",
"Xinyu",
""
],
[
"Zhou",
"Shuchang",
""
],
[
"Yao",
"Cong",
""
],
[
"Cao",
"Zhimin",
""
],
[
"Yin",
"Qi",
""
]
] | TITLE: ICDAR 2015 Text Reading in the Wild Competition
ABSTRACT: Recently, text detection and recognition in natural scenes are becoming
increasing popular in the computer vision community as well as the document
analysis community. However, majority of the existing ideas, algorithms and
systems are specifically designed for English. This technical report presents
the final results of the ICDAR 2015 Text Reading in the Wild (TRW 2015)
competition, which aims at establishing a benchmark for assessing detection and
recognition algorithms devised for both Chinese and English scripts and
providing a playground for researchers from the community. In this article, we
describe in detail the dataset, tasks, evaluation protocols and participants of
this competition, and report the performance of the participating methods.
Moreover, promising directions for future research are discussed.
|
1506.03425 | Krzysztof Choromanski | Krzysztof Choromanski and Sanjiv Kumar and Xiaofeng Liu | Fast Online Clustering with Randomized Skeleton Sets | null | null | null | null | cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a new fast online clustering algorithm that reliably recovers
arbitrary-shaped data clusters in high throughout data streams. Unlike the
existing state-of-the-art online clustering methods based on k-means or
k-medoid, it does not make any restrictive generative assumptions. In addition,
in contrast to existing nonparametric clustering techniques such as DBScan or
DenStream, it gives provable theoretical guarantees. To achieve fast
clustering, we propose to represent each cluster by a skeleton set which is
updated continuously as new data is seen. A skeleton set consists of weighted
samples from the data where weights encode local densities. The size of each
skeleton set is adapted according to the cluster geometry. The proposed
technique automatically detects the number of clusters and is robust to
outliers. The algorithm works for the infinite data stream where more than one
pass over the data is not feasible. We provide theoretical guarantees on the
quality of the clustering and also demonstrate its advantage over the existing
state-of-the-art on several datasets.
| [
{
"version": "v1",
"created": "Wed, 10 Jun 2015 18:41:55 GMT"
}
] | 2015-06-11T00:00:00 | [
[
"Choromanski",
"Krzysztof",
""
],
[
"Kumar",
"Sanjiv",
""
],
[
"Liu",
"Xiaofeng",
""
]
] | TITLE: Fast Online Clustering with Randomized Skeleton Sets
ABSTRACT: We present a new fast online clustering algorithm that reliably recovers
arbitrary-shaped data clusters in high throughout data streams. Unlike the
existing state-of-the-art online clustering methods based on k-means or
k-medoid, it does not make any restrictive generative assumptions. In addition,
in contrast to existing nonparametric clustering techniques such as DBScan or
DenStream, it gives provable theoretical guarantees. To achieve fast
clustering, we propose to represent each cluster by a skeleton set which is
updated continuously as new data is seen. A skeleton set consists of weighted
samples from the data where weights encode local densities. The size of each
skeleton set is adapted according to the cluster geometry. The proposed
technique automatically detects the number of clusters and is robust to
outliers. The algorithm works for the infinite data stream where more than one
pass over the data is not feasible. We provide theoretical guarantees on the
quality of the clustering and also demonstrate its advantage over the existing
state-of-the-art on several datasets.
|
1411.0292 | Alp Kucukelbir | Alp Kucukelbir, David M. Blei | Population Empirical Bayes | UAI 2015 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bayesian predictive inference analyzes a dataset to make predictions about
new observations. When a model does not match the data, predictive accuracy
suffers. We develop population empirical Bayes (POP-EB), a hierarchical
framework that explicitly models the empirical population distribution as part
of Bayesian analysis. We introduce a new concept, the latent dataset, as a
hierarchical variable and set the empirical population as its prior. This leads
to a new predictive density that mitigates model mismatch. We efficiently apply
this method to complex models by proposing a stochastic variational inference
algorithm, called bumping variational inference (BUMP-VI). We demonstrate
improved predictive accuracy over classical Bayesian inference in three models:
a linear regression model of health data, a Bayesian mixture model of natural
images, and a latent Dirichlet allocation topic model of scientific documents.
| [
{
"version": "v1",
"created": "Sun, 2 Nov 2014 18:50:14 GMT"
},
{
"version": "v2",
"created": "Mon, 8 Jun 2015 21:36:22 GMT"
}
] | 2015-06-10T00:00:00 | [
[
"Kucukelbir",
"Alp",
""
],
[
"Blei",
"David M.",
""
]
] | TITLE: Population Empirical Bayes
ABSTRACT: Bayesian predictive inference analyzes a dataset to make predictions about
new observations. When a model does not match the data, predictive accuracy
suffers. We develop population empirical Bayes (POP-EB), a hierarchical
framework that explicitly models the empirical population distribution as part
of Bayesian analysis. We introduce a new concept, the latent dataset, as a
hierarchical variable and set the empirical population as its prior. This leads
to a new predictive density that mitigates model mismatch. We efficiently apply
this method to complex models by proposing a stochastic variational inference
algorithm, called bumping variational inference (BUMP-VI). We demonstrate
improved predictive accuracy over classical Bayesian inference in three models:
a linear regression model of health data, a Bayesian mixture model of natural
images, and a latent Dirichlet allocation topic model of scientific documents.
|
1411.4280 | Jonathan Tompson | Jonathan Tompson, Ross Goroshin, Arjun Jain, Yann LeCun, Christopher
Bregler | Efficient Object Localization Using Convolutional Networks | 8 pages with 1 page of citations | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent state-of-the-art performance on human-body pose estimation has been
achieved with Deep Convolutional Networks (ConvNets). Traditional ConvNet
architectures include pooling and sub-sampling layers which reduce
computational requirements, introduce invariance and prevent over-training.
These benefits of pooling come at the cost of reduced localization accuracy. We
introduce a novel architecture which includes an efficient `position
refinement' model that is trained to estimate the joint offset location within
a small region of the image. This refinement model is jointly trained in
cascade with a state-of-the-art ConvNet model to achieve improved accuracy in
human joint location estimation. We show that the variance of our detector
approaches the variance of human annotations on the FLIC dataset and
outperforms all existing approaches on the MPII-human-pose dataset.
| [
{
"version": "v1",
"created": "Sun, 16 Nov 2014 17:23:02 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Apr 2015 16:55:05 GMT"
},
{
"version": "v3",
"created": "Tue, 9 Jun 2015 12:29:21 GMT"
}
] | 2015-06-10T00:00:00 | [
[
"Tompson",
"Jonathan",
""
],
[
"Goroshin",
"Ross",
""
],
[
"Jain",
"Arjun",
""
],
[
"LeCun",
"Yann",
""
],
[
"Bregler",
"Christopher",
""
]
] | TITLE: Efficient Object Localization Using Convolutional Networks
ABSTRACT: Recent state-of-the-art performance on human-body pose estimation has been
achieved with Deep Convolutional Networks (ConvNets). Traditional ConvNet
architectures include pooling and sub-sampling layers which reduce
computational requirements, introduce invariance and prevent over-training.
These benefits of pooling come at the cost of reduced localization accuracy. We
introduce a novel architecture which includes an efficient `position
refinement' model that is trained to estimate the joint offset location within
a small region of the image. This refinement model is jointly trained in
cascade with a state-of-the-art ConvNet model to achieve improved accuracy in
human joint location estimation. We show that the variance of our detector
approaches the variance of human annotations on the FLIC dataset and
outperforms all existing approaches on the MPII-human-pose dataset.
|
1502.04843 | Brijnesh Jain | Brijnesh Jain | Generalized Gradient Learning on Time Series under Elastic
Transformations | accepted for publication in Machine Learning | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The majority of machine learning algorithms assumes that objects are
represented as vectors. But often the objects we want to learn on are more
naturally represented by other data structures such as sequences and time
series. For these representations many standard learning algorithms are
unavailable. We generalize gradient-based learning algorithms to time series
under dynamic time warping. To this end, we introduce elastic functions, which
extend functions on time series to matrix spaces. Necessary conditions are
presented under which generalized gradient learning on time series is
consistent. We indicate how results carry over to arbitrary elastic distance
functions and to sequences consisting of symbolic elements. Specifically, four
linear classifiers are extended to time series under dynamic time warping and
applied to benchmark datasets. Results indicate that generalized gradient
learning via elastic functions have the potential to complement the
state-of-the-art in statistical pattern recognition on time series.
| [
{
"version": "v1",
"created": "Tue, 17 Feb 2015 10:08:48 GMT"
},
{
"version": "v2",
"created": "Tue, 9 Jun 2015 10:50:41 GMT"
}
] | 2015-06-10T00:00:00 | [
[
"Jain",
"Brijnesh",
""
]
] | TITLE: Generalized Gradient Learning on Time Series under Elastic
Transformations
ABSTRACT: The majority of machine learning algorithms assumes that objects are
represented as vectors. But often the objects we want to learn on are more
naturally represented by other data structures such as sequences and time
series. For these representations many standard learning algorithms are
unavailable. We generalize gradient-based learning algorithms to time series
under dynamic time warping. To this end, we introduce elastic functions, which
extend functions on time series to matrix spaces. Necessary conditions are
presented under which generalized gradient learning on time series is
consistent. We indicate how results carry over to arbitrary elastic distance
functions and to sequences consisting of symbolic elements. Specifically, four
linear classifiers are extended to time series under dynamic time warping and
applied to benchmark datasets. Results indicate that generalized gradient
learning via elastic functions have the potential to complement the
state-of-the-art in statistical pattern recognition on time series.
|
1506.02732 | Zhiguang Wang | Wei Song, Zhiguang Wang, Yangdong Ye, Ming Fan | Empirical Studies on Symbolic Aggregation Approximation Under
Statistical Perspectives for Knowledge Discovery in Time Series | 7 pages, 6 figures. Accepted by FSKD 2015 | null | null | null | cs.LG cs.IT math.IT | http://creativecommons.org/licenses/by/3.0/ | Symbolic Aggregation approXimation (SAX) has been the de facto standard
representation methods for knowledge discovery in time series on a number of
tasks and applications. So far, very little work has been done in empirically
investigating the intrinsic properties and statistical mechanics in SAX words.
In this paper, we applied several statistical measurements and proposed a new
statistical measurement, i.e. information embedding cost (IEC) to analyze the
statistical behaviors of the symbolic dynamics. Our experiments on the
benchmark datasets and the clinical signals demonstrate that SAX can always
reduce the complexity while preserving the core information embedded in the
original time series with significant embedding efficiency. Our proposed IEC
score provide a priori to determine if SAX is adequate for specific dataset,
which can be generalized to evaluate other symbolic representations. Our work
provides an analytical framework with several statistical tools to analyze,
evaluate and further improve the symbolic dynamics for knowledge discovery in
time series.
| [
{
"version": "v1",
"created": "Mon, 8 Jun 2015 23:52:04 GMT"
}
] | 2015-06-10T00:00:00 | [
[
"Song",
"Wei",
""
],
[
"Wang",
"Zhiguang",
""
],
[
"Ye",
"Yangdong",
""
],
[
"Fan",
"Ming",
""
]
] | TITLE: Empirical Studies on Symbolic Aggregation Approximation Under
Statistical Perspectives for Knowledge Discovery in Time Series
ABSTRACT: Symbolic Aggregation approXimation (SAX) has been the de facto standard
representation methods for knowledge discovery in time series on a number of
tasks and applications. So far, very little work has been done in empirically
investigating the intrinsic properties and statistical mechanics in SAX words.
In this paper, we applied several statistical measurements and proposed a new
statistical measurement, i.e. information embedding cost (IEC) to analyze the
statistical behaviors of the symbolic dynamics. Our experiments on the
benchmark datasets and the clinical signals demonstrate that SAX can always
reduce the complexity while preserving the core information embedded in the
original time series with significant embedding efficiency. Our proposed IEC
score provide a priori to determine if SAX is adequate for specific dataset,
which can be generalized to evaluate other symbolic representations. Our work
provides an analytical framework with several statistical tools to analyze,
evaluate and further improve the symbolic dynamics for knowledge discovery in
time series.
|
1506.02833 | Albert Saa-Garriga | Albert Sa\`a-Garriga and David Castells-Rufas and Jordi Carrabina | OMP2HMPP: Compiler Framework for Energy Performance Trade-off Analysis
of Automatically Generated Codes | null | IJCSI International Journal of Computer Science Issues, Volume 12,
Issue 2, March 2015 | null | null | cs.DC cs.PF cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present OMP2HMPP, a tool that, in a first step, automatically translates
OpenMP code into various possible transformations of HMPP. In a second step
OMP2HMPP executes all variants to obtain the performance and power consumption
of each transformation. The resulting trade-off can be used to choose the more
convenient version. After running the tool on a set of codes from the Polybench
benchmark we show that the best automatic transformation is equivalent to a
manual one done by an expert. Compared with original OpenMP code running in 2
quad-core processors we obtain an average speed-up of 31x and 5.86x factor in
operations per watt.
| [
{
"version": "v1",
"created": "Tue, 9 Jun 2015 09:24:38 GMT"
}
] | 2015-06-10T00:00:00 | [
[
"Saà-Garriga",
"Albert",
""
],
[
"Castells-Rufas",
"David",
""
],
[
"Carrabina",
"Jordi",
""
]
] | TITLE: OMP2HMPP: Compiler Framework for Energy Performance Trade-off Analysis
of Automatically Generated Codes
ABSTRACT: We present OMP2HMPP, a tool that, in a first step, automatically translates
OpenMP code into various possible transformations of HMPP. In a second step
OMP2HMPP executes all variants to obtain the performance and power consumption
of each transformation. The resulting trade-off can be used to choose the more
convenient version. After running the tool on a set of codes from the Polybench
benchmark we show that the best automatic transformation is equivalent to a
manual one done by an expert. Compared with original OpenMP code running in 2
quad-core processors we obtain an average speed-up of 31x and 5.86x factor in
operations per watt.
|
1202.6384 | Arthur Szlam | Arthur Szlam, Karol Gregor, Yann LeCun | Fast approximations to structured sparse coding and applications to
object classification | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a method for fast approximation of sparse coding. The input space
is subdivided by a binary decision tree, and we simultaneously learn a
dictionary and assignment of allowed dictionary elements for each leaf of the
tree. We store a lookup table with the assignments and the pseudoinverses for
each node, allowing for very fast inference. We give an algorithm for learning
the tree, the dictionary and the dictionary element assignment, and In the
process of describing this algorithm, we discuss the more general problem of
learning the groups in group structured sparse modelling. We show that our
method creates good sparse representations by using it in the object
recognition framework of \cite{lazebnik06,yang-cvpr-09}. Implementing our own
fast version of the SIFT descriptor the whole system runs at 20 frames per
second on $321 \times 481$ sized images on a laptop with a quad-core cpu, while
sacrificing very little accuracy on the Caltech 101 and 15 scenes benchmarks.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2012 21:27:14 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Szlam",
"Arthur",
""
],
[
"Gregor",
"Karol",
""
],
[
"LeCun",
"Yann",
""
]
] | TITLE: Fast approximations to structured sparse coding and applications to
object classification
ABSTRACT: We describe a method for fast approximation of sparse coding. The input space
is subdivided by a binary decision tree, and we simultaneously learn a
dictionary and assignment of allowed dictionary elements for each leaf of the
tree. We store a lookup table with the assignments and the pseudoinverses for
each node, allowing for very fast inference. We give an algorithm for learning
the tree, the dictionary and the dictionary element assignment, and In the
process of describing this algorithm, we discuss the more general problem of
learning the groups in group structured sparse modelling. We show that our
method creates good sparse representations by using it in the object
recognition framework of \cite{lazebnik06,yang-cvpr-09}. Implementing our own
fast version of the SIFT descriptor the whole system runs at 20 frames per
second on $321 \times 481$ sized images on a laptop with a quad-core cpu, while
sacrificing very little accuracy on the Caltech 101 and 15 scenes benchmarks.
|
1312.5851 | Mikael Henaff | Michael Mathieu, Mikael Henaff, Yann LeCun | Fast Training of Convolutional Networks through FFTs | null | null | null | null | cs.CV cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Convolutional networks are one of the most widely employed architectures in
computer vision and machine learning. In order to leverage their ability to
learn complex functions, large amounts of data are required for training.
Training a large convolutional network to produce state-of-the-art results can
take weeks, even when using modern GPUs. Producing labels using a trained
network can also be costly when dealing with web-scale datasets. In this work,
we present a simple algorithm which accelerates training and inference by a
significant factor, and can yield improvements of over an order of magnitude
compared to existing state-of-the-art implementations. This is done by
computing convolutions as pointwise products in the Fourier domain while
reusing the same transformed feature map many times. The algorithm is
implemented on a GPU architecture and addresses a number of related challenges.
| [
{
"version": "v1",
"created": "Fri, 20 Dec 2013 08:42:21 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jan 2014 00:28:06 GMT"
},
{
"version": "v3",
"created": "Tue, 28 Jan 2014 01:33:21 GMT"
},
{
"version": "v4",
"created": "Tue, 18 Feb 2014 03:20:51 GMT"
},
{
"version": "v5",
"created": "Thu, 6 Mar 2014 23:27:18 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Mathieu",
"Michael",
""
],
[
"Henaff",
"Mikael",
""
],
[
"LeCun",
"Yann",
""
]
] | TITLE: Fast Training of Convolutional Networks through FFTs
ABSTRACT: Convolutional networks are one of the most widely employed architectures in
computer vision and machine learning. In order to leverage their ability to
learn complex functions, large amounts of data are required for training.
Training a large convolutional network to produce state-of-the-art results can
take weeks, even when using modern GPUs. Producing labels using a trained
network can also be costly when dealing with web-scale datasets. In this work,
we present a simple algorithm which accelerates training and inference by a
significant factor, and can yield improvements of over an order of magnitude
compared to existing state-of-the-art implementations. This is done by
computing convolutions as pointwise products in the Fourier domain while
reusing the same transformed feature map many times. The algorithm is
implemented on a GPU architecture and addresses a number of related challenges.
|
1405.6159 | Mariano Tepper | Mariano Tepper and Guillermo Sapiro | A Bi-clustering Framework for Consensus Problems | null | null | 10.1137/140967325 | null | cs.CV cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider grouping as a general characterization for problems such as
clustering, community detection in networks, and multiple parametric model
estimation. We are interested in merging solutions from different grouping
algorithms, distilling all their good qualities into a consensus solution. In
this paper, we propose a bi-clustering framework and perspective for reaching
consensus in such grouping problems. In particular, this is the first time that
the task of finding/fitting multiple parametric models to a dataset is formally
posed as a consensus problem. We highlight the equivalence of these tasks and
establish the connection with the computational Gestalt program, that seeks to
provide a psychologically-inspired detection theory for visual events. We also
present a simple but powerful bi-clustering algorithm, specially tuned to the
nature of the problem we address, though general enough to handle many
different instances inscribed within our characterization. The presentation is
accompanied with diverse and extensive experimental results in clustering,
community detection, and multiple parametric model estimation in image
processing applications.
| [
{
"version": "v1",
"created": "Wed, 30 Apr 2014 21:58:10 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jun 2014 17:44:55 GMT"
},
{
"version": "v3",
"created": "Wed, 20 Aug 2014 22:12:15 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Tepper",
"Mariano",
""
],
[
"Sapiro",
"Guillermo",
""
]
] | TITLE: A Bi-clustering Framework for Consensus Problems
ABSTRACT: We consider grouping as a general characterization for problems such as
clustering, community detection in networks, and multiple parametric model
estimation. We are interested in merging solutions from different grouping
algorithms, distilling all their good qualities into a consensus solution. In
this paper, we propose a bi-clustering framework and perspective for reaching
consensus in such grouping problems. In particular, this is the first time that
the task of finding/fitting multiple parametric models to a dataset is formally
posed as a consensus problem. We highlight the equivalence of these tasks and
establish the connection with the computational Gestalt program, that seeks to
provide a psychologically-inspired detection theory for visual events. We also
present a simple but powerful bi-clustering algorithm, specially tuned to the
nature of the problem we address, though general enough to handle many
different instances inscribed within our characterization. The presentation is
accompanied with diverse and extensive experimental results in clustering,
community detection, and multiple parametric model estimation in image
processing applications.
|
1406.1476 | Toufiq Parag | Toufiq Parag, Anirban Chakraborty, Stephen Plaza and Lou Scheffer | A Context-aware Delayed Agglomeration Framework for Electron Microscopy
Segmentation | null | PLoS ONE 10(5): e0125825, 2015 | 10.1371/journal.pone.0125825 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Electron Microscopy (EM) image (or volume) segmentation has become
significantly important in recent years as an instrument for connectomics. This
paper proposes a novel agglomerative framework for EM segmentation. In
particular, given an over-segmented image or volume, we propose a novel
framework for accurately clustering regions of the same neuron. Unlike existing
agglomerative methods, the proposed context-aware algorithm divides superpixels
(over-segmented regions) of different biological entities into different
subsets and agglomerates them separately. In addition, this paper describes a
"delayed" scheme for agglomerative clustering that postpones some of the merge
decisions, pertaining to newly formed bodies, in order to generate a more
confident boundary prediction. We report significant improvements attained by
the proposed approach in segmentation accuracy over existing standard methods
on 2D and 3D datasets.
| [
{
"version": "v1",
"created": "Thu, 5 Jun 2014 18:46:38 GMT"
},
{
"version": "v2",
"created": "Tue, 24 Jun 2014 13:06:53 GMT"
},
{
"version": "v3",
"created": "Thu, 21 Aug 2014 17:22:34 GMT"
},
{
"version": "v4",
"created": "Fri, 19 Sep 2014 19:57:10 GMT"
},
{
"version": "v5",
"created": "Mon, 23 Mar 2015 15:28:02 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Parag",
"Toufiq",
""
],
[
"Chakraborty",
"Anirban",
""
],
[
"Plaza",
"Stephen",
""
],
[
"Scheffer",
"Lou",
""
]
] | TITLE: A Context-aware Delayed Agglomeration Framework for Electron Microscopy
Segmentation
ABSTRACT: Electron Microscopy (EM) image (or volume) segmentation has become
significantly important in recent years as an instrument for connectomics. This
paper proposes a novel agglomerative framework for EM segmentation. In
particular, given an over-segmented image or volume, we propose a novel
framework for accurately clustering regions of the same neuron. Unlike existing
agglomerative methods, the proposed context-aware algorithm divides superpixels
(over-segmented regions) of different biological entities into different
subsets and agglomerates them separately. In addition, this paper describes a
"delayed" scheme for agglomerative clustering that postpones some of the merge
decisions, pertaining to newly formed bodies, in order to generate a more
confident boundary prediction. We report significant improvements attained by
the proposed approach in segmentation accuracy over existing standard methods
on 2D and 3D datasets.
|
1409.2752 | Alireza Makhzani | Alireza Makhzani, Brendan Frey | Winner-Take-All Autoencoders | null | null | null | null | cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a winner-take-all method for learning hierarchical
sparse representations in an unsupervised fashion. We first introduce
fully-connected winner-take-all autoencoders which use mini-batch statistics to
directly enforce a lifetime sparsity in the activations of the hidden units. We
then propose the convolutional winner-take-all autoencoder which combines the
benefits of convolutional architectures and autoencoders for learning
shift-invariant sparse representations. We describe a way to train
convolutional autoencoders layer by layer, where in addition to lifetime
sparsity, a spatial sparsity within each feature map is achieved using
winner-take-all activation functions. We will show that winner-take-all
autoencoders can be used to to learn deep sparse representations from the
MNIST, CIFAR-10, ImageNet, Street View House Numbers and Toronto Face datasets,
and achieve competitive classification performance.
| [
{
"version": "v1",
"created": "Tue, 9 Sep 2014 14:38:43 GMT"
},
{
"version": "v2",
"created": "Sun, 7 Jun 2015 18:28:22 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Makhzani",
"Alireza",
""
],
[
"Frey",
"Brendan",
""
]
] | TITLE: Winner-Take-All Autoencoders
ABSTRACT: In this paper, we propose a winner-take-all method for learning hierarchical
sparse representations in an unsupervised fashion. We first introduce
fully-connected winner-take-all autoencoders which use mini-batch statistics to
directly enforce a lifetime sparsity in the activations of the hidden units. We
then propose the convolutional winner-take-all autoencoder which combines the
benefits of convolutional architectures and autoencoders for learning
shift-invariant sparse representations. We describe a way to train
convolutional autoencoders layer by layer, where in addition to lifetime
sparsity, a spatial sparsity within each feature map is achieved using
winner-take-all activation functions. We will show that winner-take-all
autoencoders can be used to to learn deep sparse representations from the
MNIST, CIFAR-10, ImageNet, Street View House Numbers and Toronto Face datasets,
and achieve competitive classification performance.
|
1410.2121 | Giulio Cimini | Giulio Cimini, Tiziano Squartini, Nicol\`o Musmeci, Michelangelo
Puliga, Andrea Gabrielli, Diego Garlaschelli, Stefano Battiston, Guido
Caldarelli | Reconstructing topological properties of complex networks using the
fitness model | null | Social Informatics (series: Lec. Notes Comp. Science 8852/2015),
pp. 323-333, Springer (edited by L. M. Aiello and D. McFarland) (2015) | 10.1007/978-3-319-15168-7_41 | null | cs.SI physics.soc-ph q-fin.GN | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A major problem in the study of complex socioeconomic systems is represented
by privacy issues$-$that can put severe limitations on the amount of accessible
information, forcing to build models on the basis of incomplete knowledge. In
this paper we investigate a novel method to reconstruct global topological
properties of a complex network starting from limited information. This method
uses the knowledge of an intrinsic property of the nodes (indicated as
fitness), and the number of connections of only a limited subset of nodes, in
order to generate an ensemble of exponential random graphs that are
representative of the real systems and that can be used to estimate its
topological properties. Here we focus in particular on reconstructing the most
basic properties that are commonly used to describe a network: density of
links, assortativity, clustering. We test the method on both benchmark
synthetic networks and real economic and financial systems, finding a
remarkable robustness with respect to the number of nodes used for calibration.
The method thus represents a valuable tool for gaining insights on
privacy-protected systems.
| [
{
"version": "v1",
"created": "Wed, 8 Oct 2014 13:57:51 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Cimini",
"Giulio",
""
],
[
"Squartini",
"Tiziano",
""
],
[
"Musmeci",
"Nicolò",
""
],
[
"Puliga",
"Michelangelo",
""
],
[
"Gabrielli",
"Andrea",
""
],
[
"Garlaschelli",
"Diego",
""
],
[
"Battiston",
"Stefano",
""
],
[
"Caldarelli",
"Guido",
""
]
] | TITLE: Reconstructing topological properties of complex networks using the
fitness model
ABSTRACT: A major problem in the study of complex socioeconomic systems is represented
by privacy issues$-$that can put severe limitations on the amount of accessible
information, forcing to build models on the basis of incomplete knowledge. In
this paper we investigate a novel method to reconstruct global topological
properties of a complex network starting from limited information. This method
uses the knowledge of an intrinsic property of the nodes (indicated as
fitness), and the number of connections of only a limited subset of nodes, in
order to generate an ensemble of exponential random graphs that are
representative of the real systems and that can be used to estimate its
topological properties. Here we focus in particular on reconstructing the most
basic properties that are commonly used to describe a network: density of
links, assortativity, clustering. We test the method on both benchmark
synthetic networks and real economic and financial systems, finding a
remarkable robustness with respect to the number of nodes used for calibration.
The method thus represents a valuable tool for gaining insights on
privacy-protected systems.
|
1503.01578 | Sanghyuk Chun | Sanghyuk Chun, Yung-Kyun Noh, Jinwoo Shin | Scalable Iterative Algorithm for Robust Subspace Clustering | This paper has been withdrawn by the author due to an error in the
initialization section | null | null | null | cs.DS cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Subspace clustering (SC) is a popular method for dimensionality reduction of
high-dimensional data, where it generalizes Principal Component Analysis (PCA).
Recently, several methods have been proposed to enhance the robustness of PCA
and SC, while most of them are computationally very expensive, in particular,
for high dimensional large-scale data. In this paper, we develop much faster
iterative algorithms for SC, incorporating robustness using a {\em non-squared}
$\ell_2$-norm objective. The known implementations for optimizing the objective
would be costly due to the alternative optimization of two separate objectives:
optimal cluster-membership assignment and robust subspace selection, while the
substitution of one process to a faster surrogate can cause failure in
convergence. To address the issue, we use a simplified procedure requiring
efficient matrix-vector multiplications for subspace update instead of solving
an expensive eigenvector problem at each iteration, in addition to release
nested robust PCA loops. We prove that the proposed algorithm monotonically
converges to a local minimum with approximation guarantees, e.g., it achieves
2-approximation for the robust PCA objective. In our experiments, the proposed
algorithm is shown to converge at an order of magnitude faster than known
algorithms optimizing the same objective, and have outperforms prior subspace
clustering methods in accuracy and running time for MNIST dataset.
| [
{
"version": "v1",
"created": "Thu, 5 Mar 2015 08:54:51 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jun 2015 20:47:35 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Chun",
"Sanghyuk",
""
],
[
"Noh",
"Yung-Kyun",
""
],
[
"Shin",
"Jinwoo",
""
]
] | TITLE: Scalable Iterative Algorithm for Robust Subspace Clustering
ABSTRACT: Subspace clustering (SC) is a popular method for dimensionality reduction of
high-dimensional data, where it generalizes Principal Component Analysis (PCA).
Recently, several methods have been proposed to enhance the robustness of PCA
and SC, while most of them are computationally very expensive, in particular,
for high dimensional large-scale data. In this paper, we develop much faster
iterative algorithms for SC, incorporating robustness using a {\em non-squared}
$\ell_2$-norm objective. The known implementations for optimizing the objective
would be costly due to the alternative optimization of two separate objectives:
optimal cluster-membership assignment and robust subspace selection, while the
substitution of one process to a faster surrogate can cause failure in
convergence. To address the issue, we use a simplified procedure requiring
efficient matrix-vector multiplications for subspace update instead of solving
an expensive eigenvector problem at each iteration, in addition to release
nested robust PCA loops. We prove that the proposed algorithm monotonically
converges to a local minimum with approximation guarantees, e.g., it achieves
2-approximation for the robust PCA objective. In our experiments, the proposed
algorithm is shown to converge at an order of magnitude faster than known
algorithms optimizing the same objective, and have outperforms prior subspace
clustering methods in accuracy and running time for MNIST dataset.
|
1503.05451 | Katharina Muelling | Katharina Muelling, Arun Venkatraman, Jean-Sebastien Valois, John
Downey, Jeffrey Weiss, Shervin Javdani, Martial Hebert, Andrew B. Schwartz,
Jennifer L. Collinger, J. Andrew Bagnell | Autonomy Infused Teleoperation with Application to BCI Manipulation | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robot teleoperation systems face a common set of challenges including
latency, low-dimensional user commands, and asymmetric control inputs. User
control with Brain-Computer Interfaces (BCIs) exacerbates these problems
through especially noisy and erratic low-dimensional motion commands due to the
difficulty in decoding neural activity. We introduce a general framework to
address these challenges through a combination of computer vision, user intent
inference, and arbitration between the human input and autonomous control
schemes. Adjustable levels of assistance allow the system to balance the
operator's capabilities and feelings of comfort and control while compensating
for a task's difficulty. We present experimental results demonstrating
significant performance improvement using the shared-control assistance
framework on adapted rehabilitation benchmarks with two subjects implanted with
intracortical brain-computer interfaces controlling a seven degree-of-freedom
robotic manipulator as a prosthetic. Our results further indicate that shared
assistance mitigates perceived user difficulty and even enables successful
performance on previously infeasible tasks. We showcase the extensibility of
our architecture with applications to quality-of-life tasks such as opening a
door, pouring liquids from containers, and manipulation with novel objects in
densely cluttered environments.
| [
{
"version": "v1",
"created": "Wed, 18 Mar 2015 15:24:12 GMT"
},
{
"version": "v2",
"created": "Sun, 7 Jun 2015 17:28:59 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Muelling",
"Katharina",
""
],
[
"Venkatraman",
"Arun",
""
],
[
"Valois",
"Jean-Sebastien",
""
],
[
"Downey",
"John",
""
],
[
"Weiss",
"Jeffrey",
""
],
[
"Javdani",
"Shervin",
""
],
[
"Hebert",
"Martial",
""
],
[
"Schwartz",
"Andrew B.",
""
],
[
"Collinger",
"Jennifer L.",
""
],
[
"Bagnell",
"J. Andrew",
""
]
] | TITLE: Autonomy Infused Teleoperation with Application to BCI Manipulation
ABSTRACT: Robot teleoperation systems face a common set of challenges including
latency, low-dimensional user commands, and asymmetric control inputs. User
control with Brain-Computer Interfaces (BCIs) exacerbates these problems
through especially noisy and erratic low-dimensional motion commands due to the
difficulty in decoding neural activity. We introduce a general framework to
address these challenges through a combination of computer vision, user intent
inference, and arbitration between the human input and autonomous control
schemes. Adjustable levels of assistance allow the system to balance the
operator's capabilities and feelings of comfort and control while compensating
for a task's difficulty. We present experimental results demonstrating
significant performance improvement using the shared-control assistance
framework on adapted rehabilitation benchmarks with two subjects implanted with
intracortical brain-computer interfaces controlling a seven degree-of-freedom
robotic manipulator as a prosthetic. Our results further indicate that shared
assistance mitigates perceived user difficulty and even enables successful
performance on previously infeasible tasks. We showcase the extensibility of
our architecture with applications to quality-of-life tasks such as opening a
door, pouring liquids from containers, and manipulation with novel objects in
densely cluttered environments.
|
1506.01245 | Xinhua Zhu | Xinhua Zhu, Fei Li, Hongchao Chen, Qi Peng | A density compensation-based path computing model for measuring semantic
similarity | 17 pages,11 figures | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The shortest path between two concepts in a taxonomic ontology is commonly
used to represent the semantic distance between concepts in the edge-based
semantic similarity measures. In the past, the edge counting is considered to
be the default method for the path computation, which is simple, intuitive and
has low computational complexity. However, a large lexical taxonomy of such as
WordNet has the irregular densities of links between concepts due to its broad
domain but. The edge counting-based path computation is powerless for this
non-uniformity problem. In this paper, we advocate that the path computation is
able to be separated from the edge-based similarity measures and form various
general computing models. Therefore, in order to solve the problem of
non-uniformity of concept density in a large taxonomic ontology, we propose a
new path computing model based on the compensation of local area density of
concepts, which is equal to the number of direct hyponyms of the subsumers of
concepts in their shortest path. This path model considers the local area
density of concepts as an extension of the edge-based path and converts the
local area density divided by their depth into the compensation for edge-based
path with an adjustable parameter, which idea has been proven to be consistent
with the information theory. This model is a general path computing model and
can be applied in various edge-based similarity algorithms. The experiment
results show that the proposed path model improves the average correlation
between edge-based measures with human judgments on Miller and Charles
benchmark from less than 0.8 to more than 0.85, and has a big advantage in
efficiency than information content (IC) computation in a dynamic ontology,
thereby successfully solving the non-uniformity problem of taxonomic ontology.
| [
{
"version": "v1",
"created": "Wed, 3 Jun 2015 13:53:05 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Zhu",
"Xinhua",
""
],
[
"Li",
"Fei",
""
],
[
"Chen",
"Hongchao",
""
],
[
"Peng",
"Qi",
""
]
] | TITLE: A density compensation-based path computing model for measuring semantic
similarity
ABSTRACT: The shortest path between two concepts in a taxonomic ontology is commonly
used to represent the semantic distance between concepts in the edge-based
semantic similarity measures. In the past, the edge counting is considered to
be the default method for the path computation, which is simple, intuitive and
has low computational complexity. However, a large lexical taxonomy of such as
WordNet has the irregular densities of links between concepts due to its broad
domain but. The edge counting-based path computation is powerless for this
non-uniformity problem. In this paper, we advocate that the path computation is
able to be separated from the edge-based similarity measures and form various
general computing models. Therefore, in order to solve the problem of
non-uniformity of concept density in a large taxonomic ontology, we propose a
new path computing model based on the compensation of local area density of
concepts, which is equal to the number of direct hyponyms of the subsumers of
concepts in their shortest path. This path model considers the local area
density of concepts as an extension of the edge-based path and converts the
local area density divided by their depth into the compensation for edge-based
path with an adjustable parameter, which idea has been proven to be consistent
with the information theory. This model is a general path computing model and
can be applied in various edge-based similarity algorithms. The experiment
results show that the proposed path model improves the average correlation
between edge-based measures with human judgments on Miller and Charles
benchmark from less than 0.8 to more than 0.85, and has a big advantage in
efficiency than information content (IC) computation in a dynamic ontology,
thereby successfully solving the non-uniformity problem of taxonomic ontology.
|
1506.01744 | Kevin Chen | Chicheng Zhang, Jimin Song, Kevin C Chen, Kamalika Chaudhuri | Spectral Learning of Large Structured HMMs for Comparative Epigenomics | 27 pages, 3 figures | null | null | null | stat.ML cs.LG math.ST q-bio.GN stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop a latent variable model and an efficient spectral algorithm
motivated by the recent emergence of very large data sets of chromatin marks
from multiple human cell types. A natural model for chromatin data in one cell
type is a Hidden Markov Model (HMM); we model the relationship between multiple
cell types by connecting their hidden states by a fixed tree of known
structure. The main challenge with learning parameters of such models is that
iterative methods such as EM are very slow, while naive spectral methods result
in time and space complexity exponential in the number of cell types. We
exploit properties of the tree structure of the hidden states to provide
spectral algorithms that are more computationally efficient for current
biological datasets. We provide sample complexity bounds for our algorithm and
evaluate it experimentally on biological data from nine human cell types.
Finally, we show that beyond our specific model, some of our algorithmic ideas
can be applied to other graphical models.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 22:57:28 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Zhang",
"Chicheng",
""
],
[
"Song",
"Jimin",
""
],
[
"Chen",
"Kevin C",
""
],
[
"Chaudhuri",
"Kamalika",
""
]
] | TITLE: Spectral Learning of Large Structured HMMs for Comparative Epigenomics
ABSTRACT: We develop a latent variable model and an efficient spectral algorithm
motivated by the recent emergence of very large data sets of chromatin marks
from multiple human cell types. A natural model for chromatin data in one cell
type is a Hidden Markov Model (HMM); we model the relationship between multiple
cell types by connecting their hidden states by a fixed tree of known
structure. The main challenge with learning parameters of such models is that
iterative methods such as EM are very slow, while naive spectral methods result
in time and space complexity exponential in the number of cell types. We
exploit properties of the tree structure of the hidden states to provide
spectral algorithms that are more computationally efficient for current
biological datasets. We provide sample complexity bounds for our algorithm and
evaluate it experimentally on biological data from nine human cell types.
Finally, we show that beyond our specific model, some of our algorithmic ideas
can be applied to other graphical models.
|
1506.02075 | Antoine Bordes | Antoine Bordes, Nicolas Usunier, Sumit Chopra, Jason Weston | Large-scale Simple Question Answering with Memory Networks | null | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training large-scale question answering systems is complicated because
training sources usually cover a small portion of the range of possible
questions. This paper studies the impact of multitask and transfer learning for
simple question answering; a setting for which the reasoning required to answer
is quite easy, as long as one can retrieve the correct evidence given a
question, which can be difficult in large-scale conditions. To this end, we
introduce a new dataset of 100k questions that we use in conjunction with
existing benchmarks. We conduct our study within the framework of Memory
Networks (Weston et al., 2015) because this perspective allows us to eventually
scale up to more complex reasoning, and show that Memory Networks can be
successfully trained to achieve excellent performance.
| [
{
"version": "v1",
"created": "Fri, 5 Jun 2015 21:48:39 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Bordes",
"Antoine",
""
],
[
"Usunier",
"Nicolas",
""
],
[
"Chopra",
"Sumit",
""
],
[
"Weston",
"Jason",
""
]
] | TITLE: Large-scale Simple Question Answering with Memory Networks
ABSTRACT: Training large-scale question answering systems is complicated because
training sources usually cover a small portion of the range of possible
questions. This paper studies the impact of multitask and transfer learning for
simple question answering; a setting for which the reasoning required to answer
is quite easy, as long as one can retrieve the correct evidence given a
question, which can be difficult in large-scale conditions. To this end, we
introduce a new dataset of 100k questions that we use in conjunction with
existing benchmarks. We conduct our study within the framework of Memory
Networks (Weston et al., 2015) because this perspective allows us to eventually
scale up to more complex reasoning, and show that Memory Networks can be
successfully trained to achieve excellent performance.
|
1506.02079 | Michael Kazhdan | Michael Kazhdan, Kunal Lillaney, William Roncal, Davi Bock, Joshua
Vogelstein, and Randal Burns | Gradient-Domain Fusion for Color Correction in Large EM Image Stacks | null | null | null | null | cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a new gradient-domain technique for processing registered EM image
stacks to remove inter-image discontinuities while preserving intra-image
detail. To this end, we process the image stack by first performing anisotropic
smoothing along the slice axis and then solving a Poisson equation within each
slice to re-introduce the detail. The final image stack is continuous across
the slice axis and maintains sharp details within each slice. Adapting existing
out-of-core techniques for solving the linear system, we describe a parallel
algorithm with time complexity that is linear in the size of the data and space
complexity that is sub-linear, allowing us to process datasets as large as five
teravoxels with a 600 MB memory footprint.
| [
{
"version": "v1",
"created": "Fri, 5 Jun 2015 22:35:31 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Kazhdan",
"Michael",
""
],
[
"Lillaney",
"Kunal",
""
],
[
"Roncal",
"William",
""
],
[
"Bock",
"Davi",
""
],
[
"Vogelstein",
"Joshua",
""
],
[
"Burns",
"Randal",
""
]
] | TITLE: Gradient-Domain Fusion for Color Correction in Large EM Image Stacks
ABSTRACT: We propose a new gradient-domain technique for processing registered EM image
stacks to remove inter-image discontinuities while preserving intra-image
detail. To this end, we process the image stack by first performing anisotropic
smoothing along the slice axis and then solving a Poisson equation within each
slice to re-introduce the detail. The final image stack is continuous across
the slice axis and maintains sharp details within each slice. Adapting existing
out-of-core techniques for solving the linear system, we describe a parallel
algorithm with time complexity that is linear in the size of the data and space
complexity that is sub-linear, allowing us to process datasets as large as five
teravoxels with a 600 MB memory footprint.
|
1506.02154 | Benyuan Liu | Benyuan Liu and Hongqi Fan and Qiang Fu and Zhilin Zhang | Bayesian De-quantization and Data Compression for Low-Energy
Physiological Signal Telemonitoring | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the issue of applying quantized compressed sensing (CS) on
low-energy telemonitoring. So far, few works studied this problem in
applications where signals were only approximately sparse. We propose a
two-stage data compressor based on quantized CS, where signals are compressed
by compressed sensing and then the compressed measurements are quantized with
only 2 bits per measurement. This compressor can greatly reduce the
transmission bit-budget. To recover signals from underdetermined, quantized
measurements, we develop a Bayesian De-quantization algorithm. It can exploit
both the model of quantization errors and the correlated structure of
physiological signals to improve the quality of recovery. The proposed data
compressor and the recovery algorithm are validated on a dataset recorded on 12
subjects during fast running. Experiment results showed that an averaged 2.596
beat per minute (BPM) estimation error was achieved by jointly using compressed
sensing with 50% compression ratio and a 2-bit quantizer. The results imply
that we can effectively transmit n bits instead of n samples, which is a
substantial improvement for low-energy wireless telemonitoring.
| [
{
"version": "v1",
"created": "Sat, 6 Jun 2015 14:29:49 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Liu",
"Benyuan",
""
],
[
"Fan",
"Hongqi",
""
],
[
"Fu",
"Qiang",
""
],
[
"Zhang",
"Zhilin",
""
]
] | TITLE: Bayesian De-quantization and Data Compression for Low-Energy
Physiological Signal Telemonitoring
ABSTRACT: We address the issue of applying quantized compressed sensing (CS) on
low-energy telemonitoring. So far, few works studied this problem in
applications where signals were only approximately sparse. We propose a
two-stage data compressor based on quantized CS, where signals are compressed
by compressed sensing and then the compressed measurements are quantized with
only 2 bits per measurement. This compressor can greatly reduce the
transmission bit-budget. To recover signals from underdetermined, quantized
measurements, we develop a Bayesian De-quantization algorithm. It can exploit
both the model of quantization errors and the correlated structure of
physiological signals to improve the quality of recovery. The proposed data
compressor and the recovery algorithm are validated on a dataset recorded on 12
subjects during fast running. Experiment results showed that an averaged 2.596
beat per minute (BPM) estimation error was achieved by jointly using compressed
sensing with 50% compression ratio and a 2-bit quantizer. The results imply
that we can effectively transmit n bits instead of n samples, which is a
substantial improvement for low-energy wireless telemonitoring.
|
1506.02184 | Jun Ye | Jun Ye, Hao Hu, Kai Li, Guo-Jun Qi and Kien A. Hua | First-Take-All: Temporal Order-Preserving Hashing for 3D Action Videos | 9 pages, 11 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the prevalence of the commodity depth cameras, the new paradigm of user
interfaces based on 3D motion capturing and recognition have dramatically
changed the way of interactions between human and computers. Human action
recognition, as one of the key components in these devices, plays an important
role to guarantee the quality of user experience. Although the model-driven
methods have achieved huge success, they cannot provide a scalable solution for
efficiently storing, retrieving and recognizing actions in the large-scale
applications. These models are also vulnerable to the temporal translation and
warping, as well as the variations in motion scales and execution rates. To
address these challenges, we propose to treat the 3D human action recognition
as a video-level hashing problem and propose a novel First-Take-All (FTA)
Hashing algorithm capable of hashing the entire video into hash codes of fixed
length. We demonstrate that this FTA algorithm produces a compact
representation of the video invariant to the above mentioned variations,
through which action recognition can be solved by an efficient nearest neighbor
search by the Hamming distance between the FTA hash codes. Experiments on the
public 3D human action datasets shows that the FTA algorithm can reach a
recognition accuracy higher than 80%, with about 15 bits per frame considering
there are 65 frames per video over the datasets.
| [
{
"version": "v1",
"created": "Sat, 6 Jun 2015 19:36:11 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Ye",
"Jun",
""
],
[
"Hu",
"Hao",
""
],
[
"Li",
"Kai",
""
],
[
"Qi",
"Guo-Jun",
""
],
[
"Hua",
"Kien A.",
""
]
] | TITLE: First-Take-All: Temporal Order-Preserving Hashing for 3D Action Videos
ABSTRACT: With the prevalence of the commodity depth cameras, the new paradigm of user
interfaces based on 3D motion capturing and recognition have dramatically
changed the way of interactions between human and computers. Human action
recognition, as one of the key components in these devices, plays an important
role to guarantee the quality of user experience. Although the model-driven
methods have achieved huge success, they cannot provide a scalable solution for
efficiently storing, retrieving and recognizing actions in the large-scale
applications. These models are also vulnerable to the temporal translation and
warping, as well as the variations in motion scales and execution rates. To
address these challenges, we propose to treat the 3D human action recognition
as a video-level hashing problem and propose a novel First-Take-All (FTA)
Hashing algorithm capable of hashing the entire video into hash codes of fixed
length. We demonstrate that this FTA algorithm produces a compact
representation of the video invariant to the above mentioned variations,
through which action recognition can be solved by an efficient nearest neighbor
search by the Hamming distance between the FTA hash codes. Experiments on the
public 3D human action datasets shows that the FTA algorithm can reach a
recognition accuracy higher than 80%, with about 15 bits per frame considering
there are 65 frames per video over the datasets.
|
1506.02203 | Matteo Ruggero Ronchi | Matteo Ruggero Ronchi and Pietro Perona | Describing Common Human Visual Actions in Images | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Which common human actions and interactions are recognizable in monocular
still images? Which involve objects and/or other people? How many is a person
performing at a time? We address these questions by exploring the actions and
interactions that are detectable in the images of the MS COCO dataset. We make
two main contributions. First, a list of 140 common `visual actions', obtained
by analyzing the largest on-line verb lexicon currently available for English
(VerbNet) and human sentences used to describe images in MS COCO. Second, a
complete set of annotations for those `visual actions', composed of
subject-object and associated verb, which we call COCO-a (a for `actions').
COCO-a is larger than existing action datasets in terms of number of actions
and instances of these actions, and is unique because it is data-driven, rather
than experimenter-biased. Other unique features are that it is exhaustive, and
that all subjects and objects are localized. A statistical analysis of the
accuracy of our annotations and of each action, interaction and subject-object
combination is provided.
| [
{
"version": "v1",
"created": "Sun, 7 Jun 2015 00:33:23 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Ronchi",
"Matteo Ruggero",
""
],
[
"Perona",
"Pietro",
""
]
] | TITLE: Describing Common Human Visual Actions in Images
ABSTRACT: Which common human actions and interactions are recognizable in monocular
still images? Which involve objects and/or other people? How many is a person
performing at a time? We address these questions by exploring the actions and
interactions that are detectable in the images of the MS COCO dataset. We make
two main contributions. First, a list of 140 common `visual actions', obtained
by analyzing the largest on-line verb lexicon currently available for English
(VerbNet) and human sentences used to describe images in MS COCO. Second, a
complete set of annotations for those `visual actions', composed of
subject-object and associated verb, which we call COCO-a (a for `actions').
COCO-a is larger than existing action datasets in terms of number of actions
and instances of these actions, and is unique because it is data-driven, rather
than experimenter-biased. Other unique features are that it is exhaustive, and
that all subjects and objects are localized. A statistical analysis of the
accuracy of our annotations and of each action, interaction and subject-object
combination is provided.
|
1506.02211 | Chao Dong | Chao Dong and Ximei Zhu and Yubin Deng and Chen Change Loy and Yu Qiao | Boosting Optical Character Recognition: A Super-Resolution Approach | 5 pages, 8 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text image super-resolution is a challenging yet open research problem in the
computer vision community. In particular, low-resolution images hamper the
performance of typical optical character recognition (OCR) systems. In this
article, we summarize our entry to the ICDAR2015 Competition on Text Image
Super-Resolution. Experiments are based on the provided ICDAR2015 TextSR
dataset and the released Tesseract-OCR 3.02 system. We report that our winning
entry of text image super-resolution framework has largely improved the OCR
performance with low-resolution images used as input, reaching an OCR accuracy
score of 77.19%, which is comparable with that of using the original
high-resolution images 78.80%.
| [
{
"version": "v1",
"created": "Sun, 7 Jun 2015 02:29:45 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Dong",
"Chao",
""
],
[
"Zhu",
"Ximei",
""
],
[
"Deng",
"Yubin",
""
],
[
"Loy",
"Chen Change",
""
],
[
"Qiao",
"Yu",
""
]
] | TITLE: Boosting Optical Character Recognition: A Super-Resolution Approach
ABSTRACT: Text image super-resolution is a challenging yet open research problem in the
computer vision community. In particular, low-resolution images hamper the
performance of typical optical character recognition (OCR) systems. In this
article, we summarize our entry to the ICDAR2015 Competition on Text Image
Super-Resolution. Experiments are based on the provided ICDAR2015 TextSR
dataset and the released Tesseract-OCR 3.02 system. We report that our winning
entry of text image super-resolution framework has largely improved the OCR
performance with low-resolution images used as input, reaching an OCR accuracy
score of 77.19%, which is comparable with that of using the original
high-resolution images 78.80%.
|
1506.02268 | George Grispos | George Grispos, William Bradley Glisson and Tim Storer | Recovering Residual Forensic Data from Smartphone Interactions with
Cloud Storage Providers | null | 2015. In The Cloud Security Ecosystem, edited by Ryan Ko and
Kim-Kwang Raymond Choo, Syngress, Boston, Pages 347-382 | 10.1016/B978-0-12-801595-7.00016-1 | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There is a growing demand for cloud storage services such as Dropbox, Box,
Syncplicity and SugarSync. These public cloud storage services can store
gigabytes of corporate and personal data in remote data centres around the
world, which can then be synchronized to multiple devices. This creates an
environment which is potentially conducive to security incidents, data breaches
and other malicious activities. The forensic investigation of public cloud
environments presents a number of new challenges for the digital forensics
community. However, it is anticipated that end-devices such as smartphones,
will retain data from these cloud storage services. This research investigates
how forensic tools that are currently available to practitioners can be used to
provide a practical solution for the problems related to investigating cloud
storage environments. The research contribution is threefold. First, the
findings from this research support the idea that end-devices which have been
used to access cloud storage services can be used to provide a partial view of
the evidence stored in the cloud service. Second, the research provides a
comparison of the number of files which can be recovered from different
versions of cloud storage applications. In doing so, it also supports the idea
that amalgamating the files recovered from more than one device can result in
the recovery of a more complete dataset. Third, the chapter contributes to the
documentation and evidentiary discussion of the artefacts created from specific
cloud storage applications and different versions of these applications on iOS
and Android smartphones.
| [
{
"version": "v1",
"created": "Sun, 7 Jun 2015 14:07:12 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Grispos",
"George",
""
],
[
"Glisson",
"William Bradley",
""
],
[
"Storer",
"Tim",
""
]
] | TITLE: Recovering Residual Forensic Data from Smartphone Interactions with
Cloud Storage Providers
ABSTRACT: There is a growing demand for cloud storage services such as Dropbox, Box,
Syncplicity and SugarSync. These public cloud storage services can store
gigabytes of corporate and personal data in remote data centres around the
world, which can then be synchronized to multiple devices. This creates an
environment which is potentially conducive to security incidents, data breaches
and other malicious activities. The forensic investigation of public cloud
environments presents a number of new challenges for the digital forensics
community. However, it is anticipated that end-devices such as smartphones,
will retain data from these cloud storage services. This research investigates
how forensic tools that are currently available to practitioners can be used to
provide a practical solution for the problems related to investigating cloud
storage environments. The research contribution is threefold. First, the
findings from this research support the idea that end-devices which have been
used to access cloud storage services can be used to provide a partial view of
the evidence stored in the cloud service. Second, the research provides a
comparison of the number of files which can be recovered from different
versions of cloud storage applications. In doing so, it also supports the idea
that amalgamating the files recovered from more than one device can result in
the recovery of a more complete dataset. Third, the chapter contributes to the
documentation and evidentiary discussion of the artefacts created from specific
cloud storage applications and different versions of these applications on iOS
and Android smartphones.
|
1506.02289 | Oana Goga | Oana Goga, Patrick Loiseau, Robin Sommer, Renata Teixeira, Krishna P.
Gummadi | On the Reliability of Profile Matching Across Large Online Social
Networks | 12 pages. To appear in KDD 2015. Extended version | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Matching the profiles of a user across multiple online social networks brings
opportunities for new services and applications as well as new insights on user
online behavior, yet it raises serious privacy concerns. Prior literature has
proposed methods to match profiles and showed that it is possible to do it
accurately, but using evaluations that focused on sampled datasets only. In
this paper, we study the extent to which we can reliably match profiles in
practice, across real-world social networks, by exploiting public attributes,
i.e., information users publicly provide about themselves. Today's social
networks have hundreds of millions of users, which brings completely new
challenges as a reliable matching scheme must identify the correct matching
profile out of the millions of possible profiles. We first define a set of
properties for profile attributes--Availability, Consistency,
non-Impersonability, and Discriminability (ACID)--that are both necessary and
sufficient to determine the reliability of a matching scheme. Using these
properties, we propose a method to evaluate the accuracy of matching schemes in
real practical cases. Our results show that the accuracy in practice is
significantly lower than the one reported in prior literature. When considering
entire social networks, there is a non-negligible number of profiles that
belong to different users but have similar attributes, which leads to many
false matches. Our paper sheds light on the limits of matching profiles in the
real world and illustrates the correct methodology to evaluate matching schemes
in realistic scenarios.
| [
{
"version": "v1",
"created": "Sun, 7 Jun 2015 17:42:45 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Goga",
"Oana",
""
],
[
"Loiseau",
"Patrick",
""
],
[
"Sommer",
"Robin",
""
],
[
"Teixeira",
"Renata",
""
],
[
"Gummadi",
"Krishna P.",
""
]
] | TITLE: On the Reliability of Profile Matching Across Large Online Social
Networks
ABSTRACT: Matching the profiles of a user across multiple online social networks brings
opportunities for new services and applications as well as new insights on user
online behavior, yet it raises serious privacy concerns. Prior literature has
proposed methods to match profiles and showed that it is possible to do it
accurately, but using evaluations that focused on sampled datasets only. In
this paper, we study the extent to which we can reliably match profiles in
practice, across real-world social networks, by exploiting public attributes,
i.e., information users publicly provide about themselves. Today's social
networks have hundreds of millions of users, which brings completely new
challenges as a reliable matching scheme must identify the correct matching
profile out of the millions of possible profiles. We first define a set of
properties for profile attributes--Availability, Consistency,
non-Impersonability, and Discriminability (ACID)--that are both necessary and
sufficient to determine the reliability of a matching scheme. Using these
properties, we propose a method to evaluate the accuracy of matching schemes in
real practical cases. Our results show that the accuracy in practice is
significantly lower than the one reported in prior literature. When considering
entire social networks, there is a non-negligible number of profiles that
belong to different users but have similar attributes, which leads to many
false matches. Our paper sheds light on the limits of matching profiles in the
real world and illustrates the correct methodology to evaluate matching schemes
in realistic scenarios.
|
1506.02428 | Purushottam Kar | Kush Bhatia and Prateek Jain and Purushottam Kar | Robust Regression via Hard Thresholding | 24 pages, 3 figures | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of Robust Least Squares Regression (RLSR) where several
response variables can be adversarially corrupted. More specifically, for a
data matrix X \in R^{p x n} and an underlying model w*, the response vector is
generated as y = X'w* + b where b \in R^n is the corruption vector supported
over at most C.n coordinates. Existing exact recovery results for RLSR focus
solely on L1-penalty based convex formulations and impose relatively strict
model assumptions such as requiring the corruptions b to be selected
independently of X.
In this work, we study a simple hard-thresholding algorithm called TORRENT
which, under mild conditions on X, can recover w* exactly even if b corrupts
the response variables in an adversarial manner, i.e. both the support and
entries of b are selected adversarially after observing X and w*. Our results
hold under deterministic assumptions which are satisfied if X is sampled from
any sub-Gaussian distribution. Finally unlike existing results that apply only
to a fixed w*, generated independently of X, our results are universal and hold
for any w* \in R^p.
Next, we propose gradient descent-based extensions of TORRENT that can scale
efficiently to large scale problems, such as high dimensional sparse recovery
and prove similar recovery guarantees for these extensions. Empirically we find
TORRENT, and more so its extensions, offering significantly faster recovery
than the state-of-the-art L1 solvers. For instance, even on moderate-sized
datasets (with p = 50K) with around 40% corrupted responses, a variant of our
proposed method called TORRENT-HYB is more than 20x faster than the best L1
solver.
| [
{
"version": "v1",
"created": "Mon, 8 Jun 2015 10:13:53 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Bhatia",
"Kush",
""
],
[
"Jain",
"Prateek",
""
],
[
"Kar",
"Purushottam",
""
]
] | TITLE: Robust Regression via Hard Thresholding
ABSTRACT: We study the problem of Robust Least Squares Regression (RLSR) where several
response variables can be adversarially corrupted. More specifically, for a
data matrix X \in R^{p x n} and an underlying model w*, the response vector is
generated as y = X'w* + b where b \in R^n is the corruption vector supported
over at most C.n coordinates. Existing exact recovery results for RLSR focus
solely on L1-penalty based convex formulations and impose relatively strict
model assumptions such as requiring the corruptions b to be selected
independently of X.
In this work, we study a simple hard-thresholding algorithm called TORRENT
which, under mild conditions on X, can recover w* exactly even if b corrupts
the response variables in an adversarial manner, i.e. both the support and
entries of b are selected adversarially after observing X and w*. Our results
hold under deterministic assumptions which are satisfied if X is sampled from
any sub-Gaussian distribution. Finally unlike existing results that apply only
to a fixed w*, generated independently of X, our results are universal and hold
for any w* \in R^p.
Next, we propose gradient descent-based extensions of TORRENT that can scale
efficiently to large scale problems, such as high dimensional sparse recovery
and prove similar recovery guarantees for these extensions. Empirically we find
TORRENT, and more so its extensions, offering significantly faster recovery
than the state-of-the-art L1 solvers. For instance, even on moderate-sized
datasets (with p = 50K) with around 40% corrupted responses, a variant of our
proposed method called TORRENT-HYB is more than 20x faster than the best L1
solver.
|
1506.02509 | Lei Zhang | Lei Zhang and David Zhang | SVM and ELM: Who Wins? Object Recognition with Deep Convolutional
Features from ImageNet | 7 pages, 4 figures | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning with a convolutional neural network (CNN) has been proved to be
very effective in feature extraction and representation of images. For image
classification problems, this work aim at finding which classifier is more
competitive based on high-level deep features of images. In this report, we
have discussed the nearest neighbor, support vector machines and extreme
learning machines for image classification under deep convolutional activation
feature representation. Specifically, we adopt the benchmark object recognition
dataset from multiple sources with domain bias for evaluating different
classifiers. The deep features of the object dataset are obtained by a
well-trained CNN with five convolutional layers and three fully-connected
layers on the challenging ImageNet. Experiments demonstrate that the ELMs
outperform SVMs in cross-domain recognition tasks. In particular,
state-of-the-art results are obtained by kernel ELM which outperforms SVMs with
about 4% of the average accuracy. The features and codes are available in
http://www.escience.cn/people/lei/index.html
| [
{
"version": "v1",
"created": "Mon, 8 Jun 2015 13:58:01 GMT"
}
] | 2015-06-09T00:00:00 | [
[
"Zhang",
"Lei",
""
],
[
"Zhang",
"David",
""
]
] | TITLE: SVM and ELM: Who Wins? Object Recognition with Deep Convolutional
Features from ImageNet
ABSTRACT: Deep learning with a convolutional neural network (CNN) has been proved to be
very effective in feature extraction and representation of images. For image
classification problems, this work aim at finding which classifier is more
competitive based on high-level deep features of images. In this report, we
have discussed the nearest neighbor, support vector machines and extreme
learning machines for image classification under deep convolutional activation
feature representation. Specifically, we adopt the benchmark object recognition
dataset from multiple sources with domain bias for evaluating different
classifiers. The deep features of the object dataset are obtained by a
well-trained CNN with five convolutional layers and three fully-connected
layers on the challenging ImageNet. Experiments demonstrate that the ELMs
outperform SVMs in cross-domain recognition tasks. In particular,
state-of-the-art results are obtained by kernel ELM which outperforms SVMs with
about 4% of the average accuracy. The features and codes are available in
http://www.escience.cn/people/lei/index.html
|
1503.05694 | Vishwesh Jatala | Vishwesh Jatala, Jayvant Anantpur, Amey Karkare | Improving GPU Performance Through Resource Sharing | null | null | null | null | cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graphics Processing Units (GPUs) consisting of Streaming Multiprocessors
(SMs) achieve high throughput by running a large number of threads and context
switching among them to hide execution latencies. The number of thread blocks,
and hence the number of threads that can be launched on an SM, depends on the
resource usage--e.g. number of registers, amount of shared memory--of the
thread blocks. Since the allocation of threads to an SM is at the thread block
granularity, some of the resources may not be used up completely and hence will
be wasted.
We propose an approach that shares the resources of SM to utilize the wasted
resources by launching more thread blocks. We show the effectiveness of our
approach for two resources: register sharing, and scratchpad (shared memory)
sharing. We further propose optimizations to hide long execution latencies,
thus reducing the number of stall cycles. We implemented our approach in
GPGPU-Sim simulator and experimentally validated it on several applications
from 4 different benchmark suites: GPGPU-Sim, Rodinia, CUDA-SDK, and Parboil.
We observed that with register sharing, applications show maximum improvement
of 24%, and average improvement of 11%. With scratchpad sharing, we observed a
maximum improvement of 30% and an average improvement of 12.5%.
| [
{
"version": "v1",
"created": "Thu, 19 Mar 2015 10:21:42 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jun 2015 07:28:13 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Jatala",
"Vishwesh",
""
],
[
"Anantpur",
"Jayvant",
""
],
[
"Karkare",
"Amey",
""
]
] | TITLE: Improving GPU Performance Through Resource Sharing
ABSTRACT: Graphics Processing Units (GPUs) consisting of Streaming Multiprocessors
(SMs) achieve high throughput by running a large number of threads and context
switching among them to hide execution latencies. The number of thread blocks,
and hence the number of threads that can be launched on an SM, depends on the
resource usage--e.g. number of registers, amount of shared memory--of the
thread blocks. Since the allocation of threads to an SM is at the thread block
granularity, some of the resources may not be used up completely and hence will
be wasted.
We propose an approach that shares the resources of SM to utilize the wasted
resources by launching more thread blocks. We show the effectiveness of our
approach for two resources: register sharing, and scratchpad (shared memory)
sharing. We further propose optimizations to hide long execution latencies,
thus reducing the number of stall cycles. We implemented our approach in
GPGPU-Sim simulator and experimentally validated it on several applications
from 4 different benchmark suites: GPGPU-Sim, Rodinia, CUDA-SDK, and Parboil.
We observed that with register sharing, applications show maximum improvement
of 24%, and average improvement of 11%. With scratchpad sharing, we observed a
maximum improvement of 30% and an average improvement of 12.5%.
|
1505.06289 | Will Monroe | Angel Chang, Will Monroe, Manolis Savva, Christopher Potts,
Christopher D. Manning | Text to 3D Scene Generation with Rich Lexical Grounding | 10 pages, 7 figures, 3 tables. To appear in ACL-IJCNLP 2015 | null | null | null | cs.CL cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ability to map descriptions of scenes to 3D geometric representations has
many applications in areas such as art, education, and robotics. However, prior
work on the text to 3D scene generation task has used manually specified object
categories and language that identifies them. We introduce a dataset of 3D
scenes annotated with natural language descriptions and learn from this data
how to ground textual descriptions to physical objects. Our method successfully
grounds a variety of lexical terms to concrete referents, and we show
quantitatively that our method improves 3D scene generation over previous work
using purely rule-based methods. We evaluate the fidelity and plausibility of
3D scenes generated with our grounding approach through human judgments. To
ease evaluation on this task, we also introduce an automated metric that
strongly correlates with human judgments.
| [
{
"version": "v1",
"created": "Sat, 23 May 2015 08:32:11 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jun 2015 01:13:17 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Chang",
"Angel",
""
],
[
"Monroe",
"Will",
""
],
[
"Savva",
"Manolis",
""
],
[
"Potts",
"Christopher",
""
],
[
"Manning",
"Christopher D.",
""
]
] | TITLE: Text to 3D Scene Generation with Rich Lexical Grounding
ABSTRACT: The ability to map descriptions of scenes to 3D geometric representations has
many applications in areas such as art, education, and robotics. However, prior
work on the text to 3D scene generation task has used manually specified object
categories and language that identifies them. We introduce a dataset of 3D
scenes annotated with natural language descriptions and learn from this data
how to ground textual descriptions to physical objects. Our method successfully
grounds a variety of lexical terms to concrete referents, and we show
quantitatively that our method improves 3D scene generation over previous work
using purely rule-based methods. We evaluate the fidelity and plausibility of
3D scenes generated with our grounding approach through human judgments. To
ease evaluation on this task, we also introduce an automated metric that
strongly correlates with human judgments.
|
1506.00243 | Hui Wang | Hui Wang and Anthony TS Ho and Shujun Li | OR-Benchmark: An Open and Reconfigurable Digital Watermarking
Benchmarking Framework | null | null | null | null | cs.MM cs.CR cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Benchmarking digital watermarking algorithms is not an easy task because
different applications of digital watermarking often have very different sets
of requirements and trade-offs between conflicting requirements. While there
have been some general-purpose digital watermarking benchmarking systems
available, they normally do not support complicated benchmarking tasks and
cannot be easily reconfigured to work with different watermarking algorithms
and testing conditions. In this paper, we propose OR-Benchmark, an open and
highly reconfigurable general-purpose digital watermarking benchmarking
framework, which has the following two key features: 1) all the interfaces are
public and general enough to support all watermarking applications and
benchmarking tasks we can think of; 2) end users can easily extend the
functionalities and freely configure what watermarking algorithms are tested,
what system components are used, how the benchmarking process runs, and what
results should be produced. We implemented a prototype of this framework as a
MATLAB software package and used it to benchmark a number of digital
watermarking algorithms involving two types of watermarks for content
authentication and self-restoration purposes. The benchmarking results
demonstrated the advantages of the proposed benchmarking framework, and also
gave us some useful insights about existing image authentication and
self-restoration watermarking algorithms which are an important but less
studied topic in digital watermarking.
| [
{
"version": "v1",
"created": "Sun, 31 May 2015 14:50:26 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jun 2015 14:49:01 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Wang",
"Hui",
""
],
[
"Ho",
"Anthony TS",
""
],
[
"Li",
"Shujun",
""
]
] | TITLE: OR-Benchmark: An Open and Reconfigurable Digital Watermarking
Benchmarking Framework
ABSTRACT: Benchmarking digital watermarking algorithms is not an easy task because
different applications of digital watermarking often have very different sets
of requirements and trade-offs between conflicting requirements. While there
have been some general-purpose digital watermarking benchmarking systems
available, they normally do not support complicated benchmarking tasks and
cannot be easily reconfigured to work with different watermarking algorithms
and testing conditions. In this paper, we propose OR-Benchmark, an open and
highly reconfigurable general-purpose digital watermarking benchmarking
framework, which has the following two key features: 1) all the interfaces are
public and general enough to support all watermarking applications and
benchmarking tasks we can think of; 2) end users can easily extend the
functionalities and freely configure what watermarking algorithms are tested,
what system components are used, how the benchmarking process runs, and what
results should be produced. We implemented a prototype of this framework as a
MATLAB software package and used it to benchmark a number of digital
watermarking algorithms involving two types of watermarks for content
authentication and self-restoration purposes. The benchmarking results
demonstrated the advantages of the proposed benchmarking framework, and also
gave us some useful insights about existing image authentication and
self-restoration watermarking algorithms which are an important but less
studied topic in digital watermarking.
|
1506.01597 | Piji Li | Lidong Bing, Piji Li, Yi Liao, Wai Lam, Weiwei Guo, Rebecca J.
Passonneau | Abstractive Multi-Document Summarization via Phrase Selection and
Merging | 11 pages, 1 figure, accepted as a full paper at ACL 2015 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an abstraction-based multi-document summarization framework that
can construct new sentences by exploring more fine-grained syntactic units than
sentences, namely, noun/verb phrases. Different from existing abstraction-based
approaches, our method first constructs a pool of concepts and facts
represented by phrases from the input documents. Then new sentences are
generated by selecting and merging informative phrases to maximize the salience
of phrases and meanwhile satisfy the sentence construction constraints. We
employ integer linear optimization for conducting phrase selection and merging
simultaneously in order to achieve the global optimal solution for a summary.
Experimental results on the benchmark data set TAC 2011 show that our framework
outperforms the state-of-the-art models under automated pyramid evaluation
metric, and achieves reasonably well results on manual linguistic quality
evaluation.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 14:04:10 GMT"
},
{
"version": "v2",
"created": "Fri, 5 Jun 2015 15:02:46 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Bing",
"Lidong",
""
],
[
"Li",
"Piji",
""
],
[
"Liao",
"Yi",
""
],
[
"Lam",
"Wai",
""
],
[
"Guo",
"Weiwei",
""
],
[
"Passonneau",
"Rebecca J.",
""
]
] | TITLE: Abstractive Multi-Document Summarization via Phrase Selection and
Merging
ABSTRACT: We propose an abstraction-based multi-document summarization framework that
can construct new sentences by exploring more fine-grained syntactic units than
sentences, namely, noun/verb phrases. Different from existing abstraction-based
approaches, our method first constructs a pool of concepts and facts
represented by phrases from the input documents. Then new sentences are
generated by selecting and merging informative phrases to maximize the salience
of phrases and meanwhile satisfy the sentence construction constraints. We
employ integer linear optimization for conducting phrase selection and merging
simultaneously in order to achieve the global optimal solution for a summary.
Experimental results on the benchmark data set TAC 2011 show that our framework
outperforms the state-of-the-art models under automated pyramid evaluation
metric, and achieves reasonably well results on manual linguistic quality
evaluation.
|
1506.01829 | Remi Lajugie | R\'emi Lajugie (SIERRA, DI-ENS), Piotr Bojanowski (WILLOW, DI-ENS),
Sylvain Arlot (SIERRA, DI-ENS), Francis Bach (SIERRA, DI-ENS) | Semidefinite and Spectral Relaxations for Multi-Label Classification | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we address the problem of multi-label classification. We
consider linear classifiers and propose to learn a prior over the space of
labels to directly leverage the performance of such methods. This prior takes
the form of a quadratic function of the labels and permits to encode both
attractive and repulsive relations between labels. We cast this problem as a
structured prediction one aiming at optimizing either the accuracies of the
predictors or the F 1-score. This leads to an optimization problem closely
related to the max-cut problem, which naturally leads to semidefinite and
spectral relaxations. We show on standard datasets how such a general prior can
improve the performances of multi-label techniques.
| [
{
"version": "v1",
"created": "Fri, 5 Jun 2015 09:19:01 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Lajugie",
"Rémi",
"",
"SIERRA, DI-ENS"
],
[
"Bojanowski",
"Piotr",
"",
"WILLOW, DI-ENS"
],
[
"Arlot",
"Sylvain",
"",
"SIERRA, DI-ENS"
],
[
"Bach",
"Francis",
"",
"SIERRA, DI-ENS"
]
] | TITLE: Semidefinite and Spectral Relaxations for Multi-Label Classification
ABSTRACT: In this paper, we address the problem of multi-label classification. We
consider linear classifiers and propose to learn a prior over the space of
labels to directly leverage the performance of such methods. This prior takes
the form of a quadratic function of the labels and permits to encode both
attractive and repulsive relations between labels. We cast this problem as a
structured prediction one aiming at optimizing either the accuracies of the
predictors or the F 1-score. This leads to an optimization problem closely
related to the max-cut problem, which naturally leads to semidefinite and
spectral relaxations. We show on standard datasets how such a general prior can
improve the performances of multi-label techniques.
|
1506.02004 | Manaal Faruqui | Manaal Faruqui, Yulia Tsvetkov, Dani Yogatama, Chris Dyer, Noah Smith | Sparse Overcomplete Word Vector Representations | Proceedings of ACL 2015 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current distributed representations of words show little resemblance to
theories of lexical semantics. The former are dense and uninterpretable, the
latter largely based on familiar, discrete classes (e.g., supersenses) and
relations (e.g., synonymy and hypernymy). We propose methods that transform
word vectors into sparse (and optionally binary) vectors. The resulting
representations are more similar to the interpretable features typically used
in NLP, though they are discovered automatically from raw corpora. Because the
vectors are highly sparse, they are computationally easy to work with. Most
importantly, we find that they outperform the original vectors on benchmark
tasks.
| [
{
"version": "v1",
"created": "Fri, 5 Jun 2015 18:20:43 GMT"
}
] | 2015-06-08T00:00:00 | [
[
"Faruqui",
"Manaal",
""
],
[
"Tsvetkov",
"Yulia",
""
],
[
"Yogatama",
"Dani",
""
],
[
"Dyer",
"Chris",
""
],
[
"Smith",
"Noah",
""
]
] | TITLE: Sparse Overcomplete Word Vector Representations
ABSTRACT: Current distributed representations of words show little resemblance to
theories of lexical semantics. The former are dense and uninterpretable, the
latter largely based on familiar, discrete classes (e.g., supersenses) and
relations (e.g., synonymy and hypernymy). We propose methods that transform
word vectors into sparse (and optionally binary) vectors. The resulting
representations are more similar to the interpretable features typically used
in NLP, though they are discovered automatically from raw corpora. Because the
vectors are highly sparse, they are computationally easy to work with. Most
importantly, we find that they outperform the original vectors on benchmark
tasks.
|
1205.4080 | Justin Ziniel | Justin Ziniel and Philip Schniter | Dynamic Compressive Sensing of Time-Varying Signals via Approximate
Message Passing | 32 pages, 7 figures | null | 10.1109/TSP.2013.2273196 | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work the dynamic compressive sensing (CS) problem of recovering
sparse, correlated, time-varying signals from sub-Nyquist, non-adaptive, linear
measurements is explored from a Bayesian perspective. While there has been a
handful of previously proposed Bayesian dynamic CS algorithms in the
literature, the ability to perform inference on high-dimensional problems in a
computationally efficient manner remains elusive. In response, we propose a
probabilistic dynamic CS signal model that captures both amplitude and support
correlation structure, and describe an approximate message passing algorithm
that performs soft signal estimation and support detection with a computational
complexity that is linear in all problem dimensions. The algorithm, DCS-AMP,
can perform either causal filtering or non-causal smoothing, and is capable of
learning model parameters adaptively from the data through an
expectation-maximization learning procedure. We provide numerical evidence that
DCS-AMP performs within 3 dB of oracle bounds on synthetic data under a variety
of operating conditions. We further describe the result of applying DCS-AMP to
two real dynamic CS datasets, as well as a frequency estimation task, to
bolster our claim that DCS-AMP is capable of offering state-of-the-art
performance and speed on real-world high-dimensional problems.
| [
{
"version": "v1",
"created": "Fri, 18 May 2012 05:33:20 GMT"
},
{
"version": "v2",
"created": "Wed, 17 Apr 2013 18:08:42 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Ziniel",
"Justin",
""
],
[
"Schniter",
"Philip",
""
]
] | TITLE: Dynamic Compressive Sensing of Time-Varying Signals via Approximate
Message Passing
ABSTRACT: In this work the dynamic compressive sensing (CS) problem of recovering
sparse, correlated, time-varying signals from sub-Nyquist, non-adaptive, linear
measurements is explored from a Bayesian perspective. While there has been a
handful of previously proposed Bayesian dynamic CS algorithms in the
literature, the ability to perform inference on high-dimensional problems in a
computationally efficient manner remains elusive. In response, we propose a
probabilistic dynamic CS signal model that captures both amplitude and support
correlation structure, and describe an approximate message passing algorithm
that performs soft signal estimation and support detection with a computational
complexity that is linear in all problem dimensions. The algorithm, DCS-AMP,
can perform either causal filtering or non-causal smoothing, and is capable of
learning model parameters adaptively from the data through an
expectation-maximization learning procedure. We provide numerical evidence that
DCS-AMP performs within 3 dB of oracle bounds on synthetic data under a variety
of operating conditions. We further describe the result of applying DCS-AMP to
two real dynamic CS datasets, as well as a frequency estimation task, to
bolster our claim that DCS-AMP is capable of offering state-of-the-art
performance and speed on real-world high-dimensional problems.
|
1206.1313 | Jos\'e Ignacio Alvarez-Hamelin Phd. | Mariano G. Beir\'o, Jorge R. Busch, Sebastian P. Grynberg and J.
Ignacio Alvarez-Hamelin | Obtaining Communities with a Fitness Growth Process | null | null | 10.1016/j.physa.2013.01.031 | null | physics.soc-ph cond-mat.stat-mech cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The study of community structure has been a hot topic of research over the
last years. But, while successfully applied in several areas, the concept lacks
of a general and precise notion. Facts like the hierarchical structure and
heterogeneity of complex networks make it difficult to unify the idea of
community and its evaluation. The global functional known as modularity is
probably the most used technique in this area. Nevertheless, its limits have
been deeply studied. Local techniques as the ones by Lancichinetti et al. and
Palla et al. arose as an answer to the resolution limit and degeneracies that
modularity has.
Here we start from the algorithm by Lancichinetti et al. and propose a unique
growth process for a fitness function that, while being local, finds a
community partition that covers the whole network, updating the scale parameter
dynamically. We test the quality of our results by using a set of benchmarks of
heterogeneous graphs. We discuss alternative measures for evaluating the
community structure and, in the light of them, infer possible explanations for
the better performance of local methods compared to global ones in these cases.
| [
{
"version": "v1",
"created": "Wed, 6 Jun 2012 19:39:46 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Beiró",
"Mariano G.",
""
],
[
"Busch",
"Jorge R.",
""
],
[
"Grynberg",
"Sebastian P.",
""
],
[
"Alvarez-Hamelin",
"J. Ignacio",
""
]
] | TITLE: Obtaining Communities with a Fitness Growth Process
ABSTRACT: The study of community structure has been a hot topic of research over the
last years. But, while successfully applied in several areas, the concept lacks
of a general and precise notion. Facts like the hierarchical structure and
heterogeneity of complex networks make it difficult to unify the idea of
community and its evaluation. The global functional known as modularity is
probably the most used technique in this area. Nevertheless, its limits have
been deeply studied. Local techniques as the ones by Lancichinetti et al. and
Palla et al. arose as an answer to the resolution limit and degeneracies that
modularity has.
Here we start from the algorithm by Lancichinetti et al. and propose a unique
growth process for a fitness function that, while being local, finds a
community partition that covers the whole network, updating the scale parameter
dynamically. We test the quality of our results by using a set of benchmarks of
heterogeneous graphs. We discuss alternative measures for evaluating the
community structure and, in the light of them, infer possible explanations for
the better performance of local methods compared to global ones in these cases.
|
1207.5661 | Rong-Hua Li | Rong-Hua Li, Jeffrey Xu Yu, Xin Huang, Hong Cheng | A Framework of Algorithms: Computing the Bias and Prestige of Nodes in
Trust Networks | null | null | 10.1371/journal.pone.0050843 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A trust network is a social network in which edges represent the trust
relationship between two nodes in the network. In a trust network, a
fundamental question is how to assess and compute the bias and prestige of the
nodes, where the bias of a node measures the trustworthiness of a node and the
prestige of a node measures the importance of the node. The larger bias of a
node implies the lower trustworthiness of the node, and the larger prestige of
a node implies the higher importance of the node. In this paper, we define a
vector-valued contractive function to characterize the bias vector which
results in a rich family of bias measurements, and we propose a framework of
algorithms for computing the bias and prestige of nodes in trust networks.
Based on our framework, we develop four algorithms that can calculate the bias
and prestige of nodes effectively and robustly. The time and space complexities
of all our algorithms are linear w.r.t. the size of the graph, thus our
algorithms are scalable to handle large datasets. We evaluate our algorithms
using five real datasets. The experimental results demonstrate the
effectiveness, robustness, and scalability of our algorithms.
| [
{
"version": "v1",
"created": "Tue, 24 Jul 2012 11:36:05 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Li",
"Rong-Hua",
""
],
[
"Yu",
"Jeffrey Xu",
""
],
[
"Huang",
"Xin",
""
],
[
"Cheng",
"Hong",
""
]
] | TITLE: A Framework of Algorithms: Computing the Bias and Prestige of Nodes in
Trust Networks
ABSTRACT: A trust network is a social network in which edges represent the trust
relationship between two nodes in the network. In a trust network, a
fundamental question is how to assess and compute the bias and prestige of the
nodes, where the bias of a node measures the trustworthiness of a node and the
prestige of a node measures the importance of the node. The larger bias of a
node implies the lower trustworthiness of the node, and the larger prestige of
a node implies the higher importance of the node. In this paper, we define a
vector-valued contractive function to characterize the bias vector which
results in a rich family of bias measurements, and we propose a framework of
algorithms for computing the bias and prestige of nodes in trust networks.
Based on our framework, we develop four algorithms that can calculate the bias
and prestige of nodes effectively and robustly. The time and space complexities
of all our algorithms are linear w.r.t. the size of the graph, thus our
algorithms are scalable to handle large datasets. We evaluate our algorithms
using five real datasets. The experimental results demonstrate the
effectiveness, robustness, and scalability of our algorithms.
|
1402.0453 | Qi Qian | Qi Qian, Rong Jin, Shenghuo Zhu and Yuanqing Lin | Fine-Grained Visual Categorization via Multi-stage Metric Learning | in CVPR 2015 | null | null | null | cs.CV cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fine-grained visual categorization (FGVC) is to categorize objects into
subordinate classes instead of basic classes. One major challenge in FGVC is
the co-occurrence of two issues: 1) many subordinate classes are highly
correlated and are difficult to distinguish, and 2) there exists the large
intra-class variation (e.g., due to object pose). This paper proposes to
explicitly address the above two issues via distance metric learning (DML). DML
addresses the first issue by learning an embedding so that data points from the
same class will be pulled together while those from different classes should be
pushed apart from each other; and it addresses the second issue by allowing the
flexibility that only a portion of the neighbors (not all data points) from the
same class need to be pulled together. However, feature representation of an
image is often high dimensional, and DML is known to have difficulty in dealing
with high dimensional feature vectors since it would require $\mathcal{O}(d^2)$
for storage and $\mathcal{O}(d^3)$ for optimization. To this end, we proposed a
multi-stage metric learning framework that divides the large-scale high
dimensional learning problem to a series of simple subproblems, achieving
$\mathcal{O}(d)$ computational complexity. The empirical study with FVGC
benchmark datasets verifies that our method is both effective and efficient
compared to the state-of-the-art FGVC approaches.
| [
{
"version": "v1",
"created": "Mon, 3 Feb 2014 18:20:53 GMT"
},
{
"version": "v2",
"created": "Thu, 4 Jun 2015 17:28:51 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Qian",
"Qi",
""
],
[
"Jin",
"Rong",
""
],
[
"Zhu",
"Shenghuo",
""
],
[
"Lin",
"Yuanqing",
""
]
] | TITLE: Fine-Grained Visual Categorization via Multi-stage Metric Learning
ABSTRACT: Fine-grained visual categorization (FGVC) is to categorize objects into
subordinate classes instead of basic classes. One major challenge in FGVC is
the co-occurrence of two issues: 1) many subordinate classes are highly
correlated and are difficult to distinguish, and 2) there exists the large
intra-class variation (e.g., due to object pose). This paper proposes to
explicitly address the above two issues via distance metric learning (DML). DML
addresses the first issue by learning an embedding so that data points from the
same class will be pulled together while those from different classes should be
pushed apart from each other; and it addresses the second issue by allowing the
flexibility that only a portion of the neighbors (not all data points) from the
same class need to be pulled together. However, feature representation of an
image is often high dimensional, and DML is known to have difficulty in dealing
with high dimensional feature vectors since it would require $\mathcal{O}(d^2)$
for storage and $\mathcal{O}(d^3)$ for optimization. To this end, we proposed a
multi-stage metric learning framework that divides the large-scale high
dimensional learning problem to a series of simple subproblems, achieving
$\mathcal{O}(d)$ computational complexity. The empirical study with FVGC
benchmark datasets verifies that our method is both effective and efficient
compared to the state-of-the-art FGVC approaches.
|
1406.4173 | D\'ora Erd\H{o}s | Dora Erdos, Vatche Ishakian, Azer Bestavros, Evimaria Terzi | A Divide-and-Conquer Algorithm for Betweenness Centrality | Shorter version of this paper appeared in Siam Data Mining 2015 | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The problem of efficiently computing the betweenness centrality of nodes has
been researched extensively. To date, the best known exact and centralized
algorithm for this task is an algorithm proposed in 2001 by Brandes. The
contribution of our paper is Brandes++, an algorithm for exact efficient
computation of betweenness centrality. The crux of our algorithm is that we
create a sketch of the graph, that we call the skeleton, by replacing subgraphs
with simpler graph structures. Depending on the underlying graph structure,
using this skeleton and by keeping appropriate summaries Brandes++ we can
achieve significantly low running times in our computations. Extensive
experimental evaluation on real life datasets demonstrate the efficacy of our
algorithm for different types of graphs. We release our code for benefit of the
research community.
| [
{
"version": "v1",
"created": "Mon, 16 Jun 2014 21:18:51 GMT"
},
{
"version": "v2",
"created": "Thu, 4 Jun 2015 19:58:34 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Erdos",
"Dora",
""
],
[
"Ishakian",
"Vatche",
""
],
[
"Bestavros",
"Azer",
""
],
[
"Terzi",
"Evimaria",
""
]
] | TITLE: A Divide-and-Conquer Algorithm for Betweenness Centrality
ABSTRACT: The problem of efficiently computing the betweenness centrality of nodes has
been researched extensively. To date, the best known exact and centralized
algorithm for this task is an algorithm proposed in 2001 by Brandes. The
contribution of our paper is Brandes++, an algorithm for exact efficient
computation of betweenness centrality. The crux of our algorithm is that we
create a sketch of the graph, that we call the skeleton, by replacing subgraphs
with simpler graph structures. Depending on the underlying graph structure,
using this skeleton and by keeping appropriate summaries Brandes++ we can
achieve significantly low running times in our computations. Extensive
experimental evaluation on real life datasets demonstrate the efficacy of our
algorithm for different types of graphs. We release our code for benefit of the
research community.
|
1504.02824 | Yelong Shen | Yelong Shen, Ruoming Jin, Jianshu Chen, Xiaodong He, Jianfeng Gao, Li
Deng | A Deep Embedding Model for Co-occurrence Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Co-occurrence Data is a common and important information source in many
areas, such as the word co-occurrence in the sentences, friends co-occurrence
in social networks and products co-occurrence in commercial transaction data,
etc, which contains rich correlation and clustering information about the
items. In this paper, we study co-occurrence data using a general energy-based
probabilistic model, and we analyze three different categories of energy-based
model, namely, the $L_1$, $L_2$ and $L_k$ models, which are able to capture
different levels of dependency in the co-occurrence data. We also discuss how
several typical existing models are related to these three types of energy
models, including the Fully Visible Boltzmann Machine (FVBM) ($L_2$), Matrix
Factorization ($L_2$), Log-BiLinear (LBL) models ($L_2$), and the Restricted
Boltzmann Machine (RBM) model ($L_k$). Then, we propose a Deep Embedding Model
(DEM) (an $L_k$ model) from the energy model in a \emph{principled} manner.
Furthermore, motivated by the observation that the partition function in the
energy model is intractable and the fact that the major objective of modeling
the co-occurrence data is to predict using the conditional probability, we
apply the \emph{maximum pseudo-likelihood} method to learn DEM. In consequence,
the developed model and its learning method naturally avoid the above
difficulties and can be easily used to compute the conditional probability in
prediction. Interestingly, our method is equivalent to learning a special
structured deep neural network using back-propagation and a special sampling
strategy, which makes it scalable on large-scale datasets. Finally, in the
experiments, we show that the DEM can achieve comparable or better results than
state-of-the-art methods on datasets across several application domains.
| [
{
"version": "v1",
"created": "Sat, 11 Apr 2015 02:56:01 GMT"
},
{
"version": "v2",
"created": "Thu, 4 Jun 2015 09:07:13 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Shen",
"Yelong",
""
],
[
"Jin",
"Ruoming",
""
],
[
"Chen",
"Jianshu",
""
],
[
"He",
"Xiaodong",
""
],
[
"Gao",
"Jianfeng",
""
],
[
"Deng",
"Li",
""
]
] | TITLE: A Deep Embedding Model for Co-occurrence Learning
ABSTRACT: Co-occurrence Data is a common and important information source in many
areas, such as the word co-occurrence in the sentences, friends co-occurrence
in social networks and products co-occurrence in commercial transaction data,
etc, which contains rich correlation and clustering information about the
items. In this paper, we study co-occurrence data using a general energy-based
probabilistic model, and we analyze three different categories of energy-based
model, namely, the $L_1$, $L_2$ and $L_k$ models, which are able to capture
different levels of dependency in the co-occurrence data. We also discuss how
several typical existing models are related to these three types of energy
models, including the Fully Visible Boltzmann Machine (FVBM) ($L_2$), Matrix
Factorization ($L_2$), Log-BiLinear (LBL) models ($L_2$), and the Restricted
Boltzmann Machine (RBM) model ($L_k$). Then, we propose a Deep Embedding Model
(DEM) (an $L_k$ model) from the energy model in a \emph{principled} manner.
Furthermore, motivated by the observation that the partition function in the
energy model is intractable and the fact that the major objective of modeling
the co-occurrence data is to predict using the conditional probability, we
apply the \emph{maximum pseudo-likelihood} method to learn DEM. In consequence,
the developed model and its learning method naturally avoid the above
difficulties and can be easily used to compute the conditional probability in
prediction. Interestingly, our method is equivalent to learning a special
structured deep neural network using back-propagation and a special sampling
strategy, which makes it scalable on large-scale datasets. Finally, in the
experiments, we show that the DEM can achieve comparable or better results than
state-of-the-art methods on datasets across several application domains.
|
1505.01861 | Tao Mei | Yingwei Pan, Tao Mei, Ting Yao, Houqiang Li, Yong Rui | Jointly Modeling Embedding and Translation to Bridge Video and Language | null | null | null | null | cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatically describing video content with natural language is a fundamental
challenge of multimedia. Recurrent Neural Networks (RNN), which models sequence
dynamics, has attracted increasing attention on visual interpretation. However,
most existing approaches generate a word locally with given previous words and
the visual content, while the relationship between sentence semantics and
visual content is not holistically exploited. As a result, the generated
sentences may be contextually correct but the semantics (e.g., subjects, verbs
or objects) are not true.
This paper presents a novel unified framework, named Long Short-Term Memory
with visual-semantic Embedding (LSTM-E), which can simultaneously explore the
learning of LSTM and visual-semantic embedding. The former aims to locally
maximize the probability of generating the next word given previous words and
visual content, while the latter is to create a visual-semantic embedding space
for enforcing the relationship between the semantics of the entire sentence and
visual content. Our proposed LSTM-E consists of three components: a 2-D and/or
3-D deep convolutional neural networks for learning powerful video
representation, a deep RNN for generating sentences, and a joint embedding
model for exploring the relationships between visual content and sentence
semantics. The experiments on YouTube2Text dataset show that our proposed
LSTM-E achieves to-date the best reported performance in generating natural
sentences: 45.3% and 31.0% in terms of BLEU@4 and METEOR, respectively. We also
demonstrate that LSTM-E is superior in predicting Subject-Verb-Object (SVO)
triplets to several state-of-the-art techniques.
| [
{
"version": "v1",
"created": "Thu, 7 May 2015 20:13:33 GMT"
},
{
"version": "v2",
"created": "Sat, 30 May 2015 10:05:50 GMT"
},
{
"version": "v3",
"created": "Thu, 4 Jun 2015 07:17:06 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Pan",
"Yingwei",
""
],
[
"Mei",
"Tao",
""
],
[
"Yao",
"Ting",
""
],
[
"Li",
"Houqiang",
""
],
[
"Rui",
"Yong",
""
]
] | TITLE: Jointly Modeling Embedding and Translation to Bridge Video and Language
ABSTRACT: Automatically describing video content with natural language is a fundamental
challenge of multimedia. Recurrent Neural Networks (RNN), which models sequence
dynamics, has attracted increasing attention on visual interpretation. However,
most existing approaches generate a word locally with given previous words and
the visual content, while the relationship between sentence semantics and
visual content is not holistically exploited. As a result, the generated
sentences may be contextually correct but the semantics (e.g., subjects, verbs
or objects) are not true.
This paper presents a novel unified framework, named Long Short-Term Memory
with visual-semantic Embedding (LSTM-E), which can simultaneously explore the
learning of LSTM and visual-semantic embedding. The former aims to locally
maximize the probability of generating the next word given previous words and
visual content, while the latter is to create a visual-semantic embedding space
for enforcing the relationship between the semantics of the entire sentence and
visual content. Our proposed LSTM-E consists of three components: a 2-D and/or
3-D deep convolutional neural networks for learning powerful video
representation, a deep RNN for generating sentences, and a joint embedding
model for exploring the relationships between visual content and sentence
semantics. The experiments on YouTube2Text dataset show that our proposed
LSTM-E achieves to-date the best reported performance in generating natural
sentences: 45.3% and 31.0% in terms of BLEU@4 and METEOR, respectively. We also
demonstrate that LSTM-E is superior in predicting Subject-Verb-Object (SVO)
triplets to several state-of-the-art techniques.
|
1506.01494 | Rui Han | Rui Han, Zhen Jia, Wanling Gao, Xinhui Tian, and Lei Wang | Benchmarking Big Data Systems: State-of-the-Art and Future Directions | 9 pages, 2 figures. arXiv admin note: substantial text overlap with
arXiv:1402.5194 | null | null | null | cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The great prosperity of big data systems such as Hadoop in recent years makes
the benchmarking of these systems become crucial for both research and industry
communities. The complexity, diversity, and rapid evolution of big data systems
gives rise to various new challenges about how we design generators to produce
data with the 4V properties (i.e. volume, velocity, variety and veracity), as
well as implement application-specific but still comprehensive workloads.
However, most of the existing big data benchmarks can be described as attempts
to solve specific problems in benchmarking systems. This article investigates
the state-of-the-art in benchmarking big data systems along with the future
challenges to be addressed to realize a successful and efficient benchmark.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 07:48:37 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Han",
"Rui",
""
],
[
"Jia",
"Zhen",
""
],
[
"Gao",
"Wanling",
""
],
[
"Tian",
"Xinhui",
""
],
[
"Wang",
"Lei",
""
]
] | TITLE: Benchmarking Big Data Systems: State-of-the-Art and Future Directions
ABSTRACT: The great prosperity of big data systems such as Hadoop in recent years makes
the benchmarking of these systems become crucial for both research and industry
communities. The complexity, diversity, and rapid evolution of big data systems
gives rise to various new challenges about how we design generators to produce
data with the 4V properties (i.e. volume, velocity, variety and veracity), as
well as implement application-specific but still comprehensive workloads.
However, most of the existing big data benchmarks can be described as attempts
to solve specific problems in benchmarking systems. This article investigates
the state-of-the-art in benchmarking big data systems along with the future
challenges to be addressed to realize a successful and efficient benchmark.
|
1506.01499 | Ashish Sureka | Ashish Sureka, Ambika Tripathi, Savita Dabral | Survey Results on Threats To External Validity, Generalizability
Concerns, Data Sharing and University-Industry Collaboration in Mining
Software Repository (MSR) Research | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mining Software Repositories (MSR) is an applied and practise-oriented field
aimed at solving real problems encountered by practitioners and bringing value
to Industry. Replication of results and findings, generalizability and external
validity, University-Industry collaboration, data sharing and creation dataset
repositories are important issues in MSR research. Research consisting of
bibliometric analysis of MSR paper shows lack of University-Industry
collaboration, deficiency of studies on closed or propriety source dataset and
lack of data as well as tool sharing by researchers. We conduct a survey of
authors of past three years of MSR conference (2012, 2013 and 2014) to collect
data on their views and suggestions to address the stated concerns. We asked 20
questions from more than 100 authors and received a response from 39 authors.
Our results shows that about one-third of the respondents always make their
dataset publicly available and about one-third believe that data sharing should
be a mandatory condition for publication in MSR conferences. Our survey reveals
that more than 50% authors used solely open-source software (OSS) dataset for
their research. More than 50% of the respondents mentioned that difficulty in
sharing Industrial dataset outside the company is one of the major impediments
in University-Industry collaboration.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 08:07:29 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Sureka",
"Ashish",
""
],
[
"Tripathi",
"Ambika",
""
],
[
"Dabral",
"Savita",
""
]
] | TITLE: Survey Results on Threats To External Validity, Generalizability
Concerns, Data Sharing and University-Industry Collaboration in Mining
Software Repository (MSR) Research
ABSTRACT: Mining Software Repositories (MSR) is an applied and practise-oriented field
aimed at solving real problems encountered by practitioners and bringing value
to Industry. Replication of results and findings, generalizability and external
validity, University-Industry collaboration, data sharing and creation dataset
repositories are important issues in MSR research. Research consisting of
bibliometric analysis of MSR paper shows lack of University-Industry
collaboration, deficiency of studies on closed or propriety source dataset and
lack of data as well as tool sharing by researchers. We conduct a survey of
authors of past three years of MSR conference (2012, 2013 and 2014) to collect
data on their views and suggestions to address the stated concerns. We asked 20
questions from more than 100 authors and received a response from 39 authors.
Our results shows that about one-third of the respondents always make their
dataset publicly available and about one-third believe that data sharing should
be a mandatory condition for publication in MSR conferences. Our survey reveals
that more than 50% authors used solely open-source software (OSS) dataset for
their research. More than 50% of the respondents mentioned that difficulty in
sharing Industrial dataset outside the company is one of the major impediments
in University-Industry collaboration.
|
1506.01596 | Roozbeh Rajabi | Roozbeh Rajabi, Hassan Ghassemian | Multilayer Structured NMF for Spectral Unmixing of Hyperspectral Images | 4 pages, conference | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the challenges in hyperspectral data analysis is the presence of mixed
pixels. Mixed pixels are the result of low spatial resolution of hyperspectral
sensors. Spectral unmixing methods decompose a mixed pixel into a set of
endmembers and abundance fractions. Due to nonnegativity constraint on
abundance fraction values, NMF based methods are well suited to this problem.
In this paper multilayer NMF has been used to improve the results of NMF
methods for spectral unmixing of hyperspectral data under the linear mixing
framework. Sparseness constraint on both spectral signatures and abundance
fractions matrices are used in this paper. Evaluation of the proposed algorithm
is done using synthetic and real datasets in terms of spectral angle and
abundance angle distances. Results show that the proposed algorithm outperforms
other previously proposed methods.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 13:53:33 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Rajabi",
"Roozbeh",
""
],
[
"Ghassemian",
"Hassan",
""
]
] | TITLE: Multilayer Structured NMF for Spectral Unmixing of Hyperspectral Images
ABSTRACT: One of the challenges in hyperspectral data analysis is the presence of mixed
pixels. Mixed pixels are the result of low spatial resolution of hyperspectral
sensors. Spectral unmixing methods decompose a mixed pixel into a set of
endmembers and abundance fractions. Due to nonnegativity constraint on
abundance fraction values, NMF based methods are well suited to this problem.
In this paper multilayer NMF has been used to improve the results of NMF
methods for spectral unmixing of hyperspectral data under the linear mixing
framework. Sparseness constraint on both spectral signatures and abundance
fractions matrices are used in this paper. Evaluation of the proposed algorithm
is done using synthetic and real datasets in terms of spectral angle and
abundance angle distances. Results show that the proposed algorithm outperforms
other previously proposed methods.
|
1506.01698 | Anna Rohrbach | Anna Rohrbach and Marcus Rohrbach and Bernt Schiele | The Long-Short Story of Movie Description | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generating descriptions for videos has many applications including assisting
blind people and human-robot interaction. The recent advances in image
captioning as well as the release of large-scale movie description datasets
such as MPII Movie Description allow to study this task in more depth. Many of
the proposed methods for image captioning rely on pre-trained object classifier
CNNs and Long-Short Term Memory recurrent networks (LSTMs) for generating
descriptions. While image description focuses on objects, we argue that it is
important to distinguish verbs, objects, and places in the challenging setting
of movie description. In this work we show how to learn robust visual
classifiers from the weak annotations of the sentence descriptions. Based on
these visual classifiers we learn how to generate a description using an LSTM.
We explore different design choices to build and train the LSTM and achieve the
best performance to date on the challenging MPII-MD dataset. We compare and
analyze our approach and prior work along various dimensions to better
understand the key challenges of the movie description task.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 19:45:36 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Rohrbach",
"Anna",
""
],
[
"Rohrbach",
"Marcus",
""
],
[
"Schiele",
"Bernt",
""
]
] | TITLE: The Long-Short Story of Movie Description
ABSTRACT: Generating descriptions for videos has many applications including assisting
blind people and human-robot interaction. The recent advances in image
captioning as well as the release of large-scale movie description datasets
such as MPII Movie Description allow to study this task in more depth. Many of
the proposed methods for image captioning rely on pre-trained object classifier
CNNs and Long-Short Term Memory recurrent networks (LSTMs) for generating
descriptions. While image description focuses on objects, we argue that it is
important to distinguish verbs, objects, and places in the challenging setting
of movie description. In this work we show how to learn robust visual
classifiers from the weak annotations of the sentence descriptions. Based on
these visual classifiers we learn how to generate a description using an LSTM.
We explore different design choices to build and train the LSTM and achieve the
best performance to date on the challenging MPII-MD dataset. We compare and
analyze our approach and prior work along various dimensions to better
understand the key challenges of the movie description task.
|
1506.01709 | H\'ector P. Mart\'inez | Vincent E. Farrugia, H\'ector P. Mart\'inez, Georgios N. Yannakakis | The Preference Learning Toolbox | null | null | null | null | stat.ML cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Preference learning (PL) is a core area of machine learning that handles
datasets with ordinal relations. As the number of generated data of ordinal
nature is increasing, the importance and role of the PL field becomes central
within machine learning research and practice. This paper introduces an open
source, scalable, efficient and accessible preference learning toolbox that
supports the key phases of the data training process incorporating various
popular data preprocessing, feature selection and preference learning methods.
| [
{
"version": "v1",
"created": "Thu, 4 Jun 2015 19:58:56 GMT"
}
] | 2015-06-05T00:00:00 | [
[
"Farrugia",
"Vincent E.",
""
],
[
"Martínez",
"Héctor P.",
""
],
[
"Yannakakis",
"Georgios N.",
""
]
] | TITLE: The Preference Learning Toolbox
ABSTRACT: Preference learning (PL) is a core area of machine learning that handles
datasets with ordinal relations. As the number of generated data of ordinal
nature is increasing, the importance and role of the PL field becomes central
within machine learning research and practice. This paper introduces an open
source, scalable, efficient and accessible preference learning toolbox that
supports the key phases of the data training process incorporating various
popular data preprocessing, feature selection and preference learning methods.
|
1204.3010 | Christian Schneider | Christian M. Schneider, Tobias A. Kesselring, Jose S. Andrade Jr. and
Hans J. Herrmann | Optimal box-covering algorithm for fractal dimension of complex networks | 5 pages, 6 figures | null | 10.1103/PhysRevE.86.016707 | null | physics.comp-ph cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The self-similarity of complex networks is typically investigated through
computational algorithms the primary task of which is to cover the structure
with a minimal number of boxes. Here we introduce a box-covering algorithm that
not only outperforms previous ones, but also finds optimal solutions. For the
two benchmark cases tested, namely, the E. Coli and the WWW networks, our
results show that the improvement can be rather substantial, reaching up to 15%
in the case of the WWW network.
| [
{
"version": "v1",
"created": "Fri, 13 Apr 2012 14:31:47 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"Schneider",
"Christian M.",
""
],
[
"Kesselring",
"Tobias A.",
""
],
[
"Andrade",
"Jose S.",
"Jr."
],
[
"Herrmann",
"Hans J.",
""
]
] | TITLE: Optimal box-covering algorithm for fractal dimension of complex networks
ABSTRACT: The self-similarity of complex networks is typically investigated through
computational algorithms the primary task of which is to cover the structure
with a minimal number of boxes. Here we introduce a box-covering algorithm that
not only outperforms previous ones, but also finds optimal solutions. For the
two benchmark cases tested, namely, the E. Coli and the WWW networks, our
results show that the improvement can be rather substantial, reaching up to 15%
in the case of the WWW network.
|
1406.4112 | Zhenyong Fu | Zhen-Yong Fu, Tao Xiang, Shaogang Gong | Semantic Graph for Zero-Shot Learning | 9 pages, 5 figures | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/3.0/ | Zero-shot learning aims to classify visual objects without any training data
via knowledge transfer between seen and unseen classes. This is typically
achieved by exploring a semantic embedding space where the seen and unseen
classes can be related. Previous works differ in what embedding space is used
and how different classes and a test image can be related. In this paper, we
utilize the annotation-free semantic word space for the former and focus on
solving the latter issue of modeling relatedness. Specifically, in contrast to
previous work which ignores the semantic relationships between seen classes and
focus merely on those between seen and unseen classes, in this paper a novel
approach based on a semantic graph is proposed to represent the relationships
between all the seen and unseen class in a semantic word space. Based on this
semantic graph, we design a special absorbing Markov chain process, in which
each unseen class is viewed as an absorbing state. After incorporating one test
image into the semantic graph, the absorbing probabilities from the test data
to each unseen class can be effectively computed; and zero-shot classification
can be achieved by finding the class label with the highest absorbing
probability. The proposed model has a closed-form solution which is linear with
respect to the number of test images. We demonstrate the effectiveness and
computational efficiency of the proposed method over the state-of-the-arts on
the AwA (animals with attributes) dataset.
| [
{
"version": "v1",
"created": "Mon, 16 Jun 2014 19:40:52 GMT"
},
{
"version": "v2",
"created": "Wed, 3 Jun 2015 09:53:18 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"Fu",
"Zhen-Yong",
""
],
[
"Xiang",
"Tao",
""
],
[
"Gong",
"Shaogang",
""
]
] | TITLE: Semantic Graph for Zero-Shot Learning
ABSTRACT: Zero-shot learning aims to classify visual objects without any training data
via knowledge transfer between seen and unseen classes. This is typically
achieved by exploring a semantic embedding space where the seen and unseen
classes can be related. Previous works differ in what embedding space is used
and how different classes and a test image can be related. In this paper, we
utilize the annotation-free semantic word space for the former and focus on
solving the latter issue of modeling relatedness. Specifically, in contrast to
previous work which ignores the semantic relationships between seen classes and
focus merely on those between seen and unseen classes, in this paper a novel
approach based on a semantic graph is proposed to represent the relationships
between all the seen and unseen class in a semantic word space. Based on this
semantic graph, we design a special absorbing Markov chain process, in which
each unseen class is viewed as an absorbing state. After incorporating one test
image into the semantic graph, the absorbing probabilities from the test data
to each unseen class can be effectively computed; and zero-shot classification
can be achieved by finding the class label with the highest absorbing
probability. The proposed model has a closed-form solution which is linear with
respect to the number of test images. We demonstrate the effectiveness and
computational efficiency of the proposed method over the state-of-the-arts on
the AwA (animals with attributes) dataset.
|
1411.5726 | Ramakrishna Vedantam | Ramakrishna Vedantam, C. Lawrence Zitnick and Devi Parikh | CIDEr: Consensus-based Image Description Evaluation | To appear in CVPR 2015 | null | null | null | cs.CV cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatically describing an image with a sentence is a long-standing
challenge in computer vision and natural language processing. Due to recent
progress in object detection, attribute classification, action recognition,
etc., there is renewed interest in this area. However, evaluating the quality
of descriptions has proven to be challenging. We propose a novel paradigm for
evaluating image descriptions that uses human consensus. This paradigm consists
of three main parts: a new triplet-based method of collecting human annotations
to measure consensus, a new automated metric (CIDEr) that captures consensus,
and two new datasets: PASCAL-50S and ABSTRACT-50S that contain 50 sentences
describing each image. Our simple metric captures human judgment of consensus
better than existing metrics across sentences generated by various sources. We
also evaluate five state-of-the-art image description approaches using this new
protocol and provide a benchmark for future comparisons. A version of CIDEr
named CIDEr-D is available as a part of MS COCO evaluation server to enable
systematic evaluation and benchmarking.
| [
{
"version": "v1",
"created": "Thu, 20 Nov 2014 23:54:35 GMT"
},
{
"version": "v2",
"created": "Wed, 3 Jun 2015 01:42:20 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"Vedantam",
"Ramakrishna",
""
],
[
"Zitnick",
"C. Lawrence",
""
],
[
"Parikh",
"Devi",
""
]
] | TITLE: CIDEr: Consensus-based Image Description Evaluation
ABSTRACT: Automatically describing an image with a sentence is a long-standing
challenge in computer vision and natural language processing. Due to recent
progress in object detection, attribute classification, action recognition,
etc., there is renewed interest in this area. However, evaluating the quality
of descriptions has proven to be challenging. We propose a novel paradigm for
evaluating image descriptions that uses human consensus. This paradigm consists
of three main parts: a new triplet-based method of collecting human annotations
to measure consensus, a new automated metric (CIDEr) that captures consensus,
and two new datasets: PASCAL-50S and ABSTRACT-50S that contain 50 sentences
describing each image. Our simple metric captures human judgment of consensus
better than existing metrics across sentences generated by various sources. We
also evaluate five state-of-the-art image description approaches using this new
protocol and provide a benchmark for future comparisons. A version of CIDEr
named CIDEr-D is available as a part of MS COCO evaluation server to enable
systematic evaluation and benchmarking.
|
1501.07430 | Seungjin Choi | Juho Lee and Seungjin Choi | Bayesian Hierarchical Clustering with Exponential Family: Small-Variance
Asymptotics and Reducibility | 10 pages, 2 figures, AISTATS-2015 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bayesian hierarchical clustering (BHC) is an agglomerative clustering method,
where a probabilistic model is defined and its marginal likelihoods are
evaluated to decide which clusters to merge. While BHC provides a few
advantages over traditional distance-based agglomerative clustering algorithms,
successive evaluation of marginal likelihoods and careful hyperparameter tuning
are cumbersome and limit the scalability. In this paper we relax BHC into a
non-probabilistic formulation, exploring small-variance asymptotics in
conjugate-exponential models. We develop a novel clustering algorithm, referred
to as relaxed BHC (RBHC), from the asymptotic limit of the BHC model that
exhibits the scalability of distance-based agglomerative clustering algorithms
as well as the flexibility of Bayesian nonparametric models. We also
investigate the reducibility of the dissimilarity measure emerged from the
asymptotic limit of the BHC model, allowing us to use scalable algorithms such
as the nearest neighbor chain algorithm. Numerical experiments on both
synthetic and real-world datasets demonstrate the validity and high performance
of our method.
| [
{
"version": "v1",
"created": "Thu, 29 Jan 2015 12:13:01 GMT"
},
{
"version": "v2",
"created": "Wed, 3 Jun 2015 00:45:09 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"Lee",
"Juho",
""
],
[
"Choi",
"Seungjin",
""
]
] | TITLE: Bayesian Hierarchical Clustering with Exponential Family: Small-Variance
Asymptotics and Reducibility
ABSTRACT: Bayesian hierarchical clustering (BHC) is an agglomerative clustering method,
where a probabilistic model is defined and its marginal likelihoods are
evaluated to decide which clusters to merge. While BHC provides a few
advantages over traditional distance-based agglomerative clustering algorithms,
successive evaluation of marginal likelihoods and careful hyperparameter tuning
are cumbersome and limit the scalability. In this paper we relax BHC into a
non-probabilistic formulation, exploring small-variance asymptotics in
conjugate-exponential models. We develop a novel clustering algorithm, referred
to as relaxed BHC (RBHC), from the asymptotic limit of the BHC model that
exhibits the scalability of distance-based agglomerative clustering algorithms
as well as the flexibility of Bayesian nonparametric models. We also
investigate the reducibility of the dissimilarity measure emerged from the
asymptotic limit of the BHC model, allowing us to use scalable algorithms such
as the nearest neighbor chain algorithm. Numerical experiments on both
synthetic and real-world datasets demonstrate the validity and high performance
of our method.
|
1506.01077 | Saullo Haniell Galv\~ao De Oliveira | Saullo Haniell Galv\~ao de Oliveira, Rosana Veroneze, Fernando Jos\'e
Von Zuben | On bicluster aggregation and its benefits for enumerative solutions | 15 pages, will be published by Springer Verlag in the LNAI Series in
the book Advances in Data Mining | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/3.0/ | Biclustering involves the simultaneous clustering of objects and their
attributes, thus defining local two-way clustering models. Recently, efficient
algorithms were conceived to enumerate all biclusters in real-valued datasets.
In this case, the solution composes a complete set of maximal and non-redundant
biclusters. However, the ability to enumerate biclusters revealed a challenging
scenario: in noisy datasets, each true bicluster may become highly fragmented
and with a high degree of overlapping. It prevents a direct analysis of the
obtained results. To revert the fragmentation, we propose here two approaches
for properly aggregating the whole set of enumerated biclusters: one based on
single linkage and the other directly exploring the rate of overlapping. Both
proposals were compared with each other and with the actual state-of-the-art in
several experiments, and they not only significantly reduced the number of
biclusters but also consistently increased the quality of the solution.
| [
{
"version": "v1",
"created": "Tue, 2 Jun 2015 22:26:42 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"de Oliveira",
"Saullo Haniell Galvão",
""
],
[
"Veroneze",
"Rosana",
""
],
[
"Von Zuben",
"Fernando José",
""
]
] | TITLE: On bicluster aggregation and its benefits for enumerative solutions
ABSTRACT: Biclustering involves the simultaneous clustering of objects and their
attributes, thus defining local two-way clustering models. Recently, efficient
algorithms were conceived to enumerate all biclusters in real-valued datasets.
In this case, the solution composes a complete set of maximal and non-redundant
biclusters. However, the ability to enumerate biclusters revealed a challenging
scenario: in noisy datasets, each true bicluster may become highly fragmented
and with a high degree of overlapping. It prevents a direct analysis of the
obtained results. To revert the fragmentation, we propose here two approaches
for properly aggregating the whole set of enumerated biclusters: one based on
single linkage and the other directly exploring the rate of overlapping. Both
proposals were compared with each other and with the actual state-of-the-art in
several experiments, and they not only significantly reduced the number of
biclusters but also consistently increased the quality of the solution.
|
1506.01092 | Seungjin Choi | Saehoon Kim and Seungjin Choi | Bilinear Random Projections for Locality-Sensitive Binary Codes | 11 pages, 23 figures, CVPR-2015 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Locality-sensitive hashing (LSH) is a popular data-independent indexing
method for approximate similarity search, where random projections followed by
quantization hash the points from the database so as to ensure that the
probability of collision is much higher for objects that are close to each
other than for those that are far apart. Most of high-dimensional visual
descriptors for images exhibit a natural matrix structure. When visual
descriptors are represented by high-dimensional feature vectors and long binary
codes are assigned, a random projection matrix requires expensive complexities
in both space and time. In this paper we analyze a bilinear random projection
method where feature matrices are transformed to binary codes by two smaller
random projection matrices. We base our theoretical analysis on extending
Raginsky and Lazebnik's result where random Fourier features are composed with
random binary quantizers to form locality sensitive binary codes. To this end,
we answer the following two questions: (1) whether a bilinear random projection
also yields similarity-preserving binary codes; (2) whether a bilinear random
projection yields performance gain or loss, compared to a large linear
projection. Regarding the first question, we present upper and lower bounds on
the expected Hamming distance between binary codes produced by bilinear random
projections. In regards to the second question, we analyze the upper and lower
bounds on covariance between two bits of binary codes, showing that the
correlation between two bits is small. Numerical experiments on MNIST and
Flickr45K datasets confirm the validity of our method.
| [
{
"version": "v1",
"created": "Wed, 3 Jun 2015 00:30:26 GMT"
}
] | 2015-06-04T00:00:00 | [
[
"Kim",
"Saehoon",
""
],
[
"Choi",
"Seungjin",
""
]
] | TITLE: Bilinear Random Projections for Locality-Sensitive Binary Codes
ABSTRACT: Locality-sensitive hashing (LSH) is a popular data-independent indexing
method for approximate similarity search, where random projections followed by
quantization hash the points from the database so as to ensure that the
probability of collision is much higher for objects that are close to each
other than for those that are far apart. Most of high-dimensional visual
descriptors for images exhibit a natural matrix structure. When visual
descriptors are represented by high-dimensional feature vectors and long binary
codes are assigned, a random projection matrix requires expensive complexities
in both space and time. In this paper we analyze a bilinear random projection
method where feature matrices are transformed to binary codes by two smaller
random projection matrices. We base our theoretical analysis on extending
Raginsky and Lazebnik's result where random Fourier features are composed with
random binary quantizers to form locality sensitive binary codes. To this end,
we answer the following two questions: (1) whether a bilinear random projection
also yields similarity-preserving binary codes; (2) whether a bilinear random
projection yields performance gain or loss, compared to a large linear
projection. Regarding the first question, we present upper and lower bounds on
the expected Hamming distance between binary codes produced by bilinear random
projections. In regards to the second question, we analyze the upper and lower
bounds on covariance between two bits of binary codes, showing that the
correlation between two bits is small. Numerical experiments on MNIST and
Flickr45K datasets confirm the validity of our method.
|
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.