If you would like to be kept up to date with latest news from PASCAL, you can follow us on Twitter.
The Twitter feed will contain the latest Job Vacancies, Calls for Participation and latest news from the network. Alternatively you can subscribe to our RSS Feed.
If you have any other items that you would like added onto the website or our Twitter Feed, please email rebecca.martin (at) cs.ucl.ac.uk.
Dear Pascal researchers,
we are proud to announce mldata, the machine learning data set
repository at http://mldata.org.
mldata is a community website aimed at exchanging data sets. Compared
to existing sites, the emphasis lies on community. That means that
anyone can upload data, comment on existing data sets, contribute
solutions to existing data sets, discuss topics in the forum, and in
general easily interact with other users.
mldata is organized into four main types of objects:
* Data – just raw data
* Task – learning tasks defined on data sets
* Method – a machine learning method, can be applied to a Task
* Challenge – a set of Tasks defining a challenge
In principle, any kind of data can be uploaded, but mldata can parse
some data formats like ARFF, CSV, and that used libsvm and other SVM
solvers. For such data sets, more functionality is available like
automatic conversion to other data sets.
Other features include automatic evaluation of solutions for tasks
using one of a large number of already available performance measures,
but of course we’re glad to add any user contributed performance measure.
So have a look, and let us know what you think on the mldata forum!
Mikio Braun – on behalf of the mldata team.
mldata is sponsored by the Pascal2 Network of Excellence.
Featured BMI Workshop
IEEE SMC 2011: 3rd Annual Workshop on Brain-Machine Interfaces
Anchorage, Alaska
October 10-11, 2011
This workshop is co-sponsored jointly by the IEEE Systems, Man, and Cybernetics Society (SMC), Circuits and Systems Society (CAS), and the Engineering, Medicine and Biology Society (EMBS).
Important Dates
April 1, 2011: Deadline for submission of full-length papers
June 1, 2011: Acceptance/Rejection notification
July 5, 2011: Final camera-ready papers due in electronic form
For more details:
http://www.smc2011.org/technical-program/bmi-workshop
1st International Workshop on Similarity-Based Pattern Analysis and Recognition
28-30 September, 2011
Venice, Italy
http://www.dsi.unive.it/~simbad
MOTIVATIONS AND OBJECTIVES
Traditional pattern recognition techniques are intimately linked to
the notion of “feature spaces.” Adopting this view, each object is
described in terms of a vector of numerical attributes and is
therefore mapped to a point in a Euclidean (geometric) vector space so
that the distances between the points reflect the observed
(dis)similarities between the respective objects. This kind of
representation is attractive because geometric spaces offer powerful
analytical as well as computational tools that are simply not
available in other representations. Indeed, classical pattern
recognition methods are tightly related to geometrical concepts and
numerous powerful tools have been developed during the last few
decades, starting from the maximal likelihood method in the 1920’s, to
perceptrons in the 1960’s, to kernel machines in the 1990’s.
However, the geometric approach suffers from a major intrinsic
limitation, which concerns the representational power of vectorial,
feature-based descriptions. In fact, there are numerous application
domains where either it is not possible to find satisfactory features
or they are inefficient for learning purposes. This modeling
difficulty typically occurs in cases when experts cannot define
features in a straightforward way (e.g., protein descriptors vs.
alignments), when data are high dimensional (e.g., images), when
features consist of both numerical and categorical variables (e.g.,
person data, like weight, sex, eye color, etc.), and in the presence
of missing or inhomogeneous data. But, probably, this situation arises
most commonly when objects are described in terms of structural
properties, such as parts and relations between parts, as is the case
in shape recognition.
In the last few years, interest around purely similarity-based
techniques has grown considerably. For example, within the supervised
learning paradigm (where expert-labeled training data is assumed to be
available) the now famous “kernel trick” shifts the focus from the
choice of an appropriate set of features to the choice of a suitable
kernel, which is related to object similarities. However, this shift
of focus is only partial, as the classical interpretation of the
notion of a kernel is that it provides an implicit transformation of
the feature space rather than a purely similarity-based
representation. Similarly, in the unsupervised domain, there has been
an increasing interest around pairwise or even multiway algorithms,
such as spectral and graph-theoretic clustering methods, which avoid
the use of features altogether.
By departing from vector-space representations one is confronted with
the challenging problem of dealing with (dis)similarities that do not
necessarily possess the Euclidean behavior or not even obey the
requirements of a metric. The lack of the Euclidean and/or metric
properties undermines the very foundations of traditional pattern
recognition theories and algorithms, and poses totally new
theoretical/computational questions and challenges.
The workshop will mark the end of the EU FP7 Projects SIMBAD
(http://simbad-fp7.eu), which was devoted precisely to these themes,
and is a follow-up of the ICML 2010 Workshop on “Learning in
non-(geo)metric spaces” (http://www.dsi.unive.it/~icml2010lngs). Its
aim is to consolidate research efforts in this area, and to provide an
informal discussion forum for researchers and practitioners interested
in this important yet diverse subject. The discussion will revolve
around two main themes, which basically correspond to the two
fundamental questions that arise when abandoning the realm of
vectorial, feature-based representations, namely:
– How can one obtain suitable similarity information from data
representations that are more powerful than, or simply different from,
the vectorial?
– How can one use similarity information in order to perform learning
and classification tasks?
We aim at covering a wide range of problems and perspectives, from
supervised to unsupervised learning, from generative to discriminative
models, and from theoretical issues to real-world practical
applications.
Accordingly, topics of interest include (but are not limited to):
– Embedding and embeddability
– Graph spectra and spectral geometry
– Indefinite and structural kernels
– Game-theoretic models of pattern recognition
– Characterization of non-(geo)metric behaviour
– Foundational issues
– Measures of (geo)metric violations
– Learning and combining similarities
– Multiple-instance learning
– Applications
FORMAT
The workshop will feature contributed talks and posters as well as
invited presentations. We feel that the more informal the better, and
we would like to solicit open and lively discussions and exchange of
ideas from researchers with different backgrounds and perspectives.
Plenty of time will be allocated to questions, discussions, and
breaks.
We plan to get videolectures coverage.
ORGANIZATION
Program Chairs
Marcello Pelillo, University of Venice, Italy
Edwin Hancock, University of York, UK
Steering Committee
Joachim Buhmann, ETH Zurich, Switzerland
Robert Duin, Delft University of Technology, The Netherlands
Mario Figueiredo, Technical University of Lisbon, Portugal
Edwin Hancock, University of York, UK
Vittorio Murino, University of Verona, Italy
Marcello Pelillo (chair), University of Venice, Italy
Program Committee
Maria-Florina Balcan, Georgia Institute of Technology, USA
Joachim Buhmann, ETH Zurich, Switzerland
Horst Bunke, University of Bern, Switzerland
Tiberio Caetano, NICTA, Australia
Umberto Castellani, University of Verona, Italy
Luca Cazzanti, University of Washington, Seattle, USA
Nicolò Cesa-Bianchi, University of Milan, Italy
Robert Duin, Delft University of Technology, The Netherlands
Francisco Escolano, University of Alicante, Spain
Mario Figueiredo, Technical University of Lisbon, Portugal
Ana Fred, Technical University of Lisbon, Portugal
Bernard Haasdonk, University of Stuttgart, Germany
Edwin Hancock, University of York, UK
Anil Jain, Michigan State University, USA
Robert Krauthgamer, Weizmann Institute of Science, Israel
Marco Loog, Delft University of Technology, The Netherlands
Vittorio Murino, University of Verona, Italy
Elzbieta Pekalska, University of Manchester, UK
Marcello Pelillo, University of Venice, Italy
Antonio Robles-Kelly, NICTA, Australia
Volker Roth, University of Basel, Switzerland
Andrea Torsello, University of Venice, Italy
Richard Wilson, University of York, UK
Organization Committee
Samuel Rota Bulò (chair), University of Venice, Italy
Nicola Rebagliati, University of Venice, Italy
Luca Rossi, University of Venice, Italy
Teresa Scantamburlo, University of Venice, Italy
IMPORTANT DATES
Paper submission: May 15, 2011
Notifications: June 19, 2011
Camera-ready due: July 2011
Conference: September 28-30, 2011
PAPER SUBMISSION
Papers must be submitted electronically at the conference website
using the EasyChair submission system. Manuscripts should be in pdf
and formatted according to Springer’s Lecture Notes in Computer
Science (LNCS) style. Information concerning typesetting can be
obtained directly from Springer at:
http://www.springer.com/comp/lncs/authors.html.
Papers must not exceed 16 pages and should report original work.
All submitted papers will be subject to a rigorous peer-review
process. Accepted papers will appear in the workshop proceedings,
which will be published in Springer’s Lecture Notes in Computer
Science (LNCS) series.
Submission implies the willingness of at least one of the authors to
register and present the paper, if accepted.
In this workshop, we aim to bring together researchers working on inference problems
in computer vision and pattern recognition, in which the specific ‘structures’ that arise in
real applications allow for reduced complexity or increased accuracy.
Well-known examples include submodularity, sparsity, and convexity.
However, there are numerous lesser-known yet important results:
exploiting shared potentials; choosing message-passing schemes based on specific
inputs; exploiting potential functions that are ‘truncated’; exploiting topology in
bipartite, planar, or grid-like models; exploiting potential functions that factorize.
Among these ideas there is a common theme: the structure of energy functions that arise
in computer vision applications often allows for far better performance than the
pessimistic results offered by standard inference procedures.
We invite submissions in the following areas:
* Exact and approximate inference in graphical models
* Exploiting graph topology: bipartite graphs; planar graphs; grid models (etc.)
* Submodularity, sparsity, convexity
* Message passing: messages that factorize; repeated messages; message-passing
schemes (etc.)
* Higher-order potentials for image labeling
* Other types of structure: shared potentials; low-order potentials (etc.)
Submissions to other areas are also encouraged. Each accepted submission will be
included in a poster session and a ‘spotlight-style’
presentation. We also invite authors of inference code and other resources to participate
in our spotlight session.
Participants are invited to submit 4-page extended abstracts in the CVPR>2011 format by April 15. Further details may be found on the workshop webpage: http://users.cecs.anu.edu.au/~julianm/cvpr2011.html
If you have any questions or comments, or wish to have a resource added to our
webpage, please e-mail Julian McAuley (julian.mcauley(at)gmail.com).
Thanks,
Julian McAuley, Tiberio Caetano, Pushmeet Kohli, Pawan Kumar, Stephen Gould
19th European Symposium on Artificial Neural Networks,
Computational Intelligence and Machine Learning
Bruges (Belgium) – April 27-28-20, 2011
Preliminary program
The preliminary program of the ESANN 2011 conference is now available on the Web:
http://www.dice.ucl.ac.be/esann
For those of you who maintain WWW pages including lists of related machine learning and artificial neural networks sites: we would appreciate if you could add the above URL to your list; thank you very much!
For 19 years the ESANN conference has become a major event in the field of neural computation and machine learning. ESANN is a selective conference focusing on fundamental aspects of artificial neural networks, machine learning, statistical information processing and computational intelligence. Mathematical foundations, algorithms and tools, and applications are covered.
The program of the conference can be found at http://www.dice.ucl.ac.be/esann, together with practical information about the conference venue, registration, etc. Other information can be obtained by sending an e-mail to esann(at)uclouvain.be.
This PhD will investigate links between neuroscience, social insect collective behaviour, and decision theory. Theories of optimal decision-making have successfully been applied to individual-level decisions, both in explaining data from experimental subjects, and in analysing the optimal performance of neurobiologically realistic models. A key feature of many such models is that sensory evidence is compared against some internal threshold in determining what choice to make. The magnitude of this difference could be thought of as a measure of confidence in the decision; if the evidence was close to the threshold then an error may have been made and confidence in the decision should be low, but if the evidence was way above the threshold, then an error is unlikely and decision confidence should be high. Decision confidence has proved to be useful in understanding behaviour and neural activity in individuals [1]. At the same time, recent complementary work has applied optimality theory developed for neural models to the collective behaviour of social insects, such as ants and honeybees, when searching for a new potential nest site [3]. There are many similarities between the interaction patterns of social insect colonies, and neural populations in the brain [2][4]. The proposed PhD project will further extend these analogies, by examining the potential role of individual and collective decision-confidence in models of house-hunting by ant and honeybee colonies.
The successful candidate will have a background in a numerate discipline such as mathematics, computer science, or physics, ideally with some knowledge of probability, statistics and decision theory. A demonstrated interest in biology is a definite advantage. They will become part of the newly established Behavioural and Evolutionary Theory Lab at the University of Sheffield, Department of Computer Science, under the direction of Dr James Marshall. It is anticipated that there will be opportunities for interaction with empirical social insect researchers to inform and test the theory developed during the project.
About the Behavioural and Evolutionary Theory Lab
—-
The Behavioural and Evolutionary Theory Lab is an interdisciplinary collection of individuals interested in how and why behaviours evolve. We are interested in behaviours and behavioural mechanisms, and their evolutionary function. We apply a range of theoretical approaches, from mathematics and statistics, decision theory, computer science, and physics. Particular topics of interest are currently the evolution of social behaviour, such as altruism and cooperation, and optimal decision-making mechanisms in groups, such as social insects, and in individuals. The Lab is part of the Department of Computer Science, University of Sheffield, and is physically based in the interdisciplinary Kroto Research Institute.
Applications are invited from UK home students and EU citizens. Fees and a stipend will be paid for the duration of the studentship. Apply online or contact Dr Marshall if you require further information. Closing date: Feb 27th.
References
—
[1] Kepecs, A. et al. (2008) Neural correlates, computation and behavioural impact of decision confidence. Nature 455, 227-231.
[2] Lindhart, E. (2009) Ants and neurons. SEED Magazine (http://seedmagazine.com/content/article/ants_and_neurons/)
[3] Marshall, J.A.R. et al. (2009) On optimal decision-making in brains and social insect colonies. Journal of the Royal Society: Interface 6, 1065-1074.
[4] Marshall, J.A.R. and Franks, N.R. (2009) Colony-level cognition. Current Biology 19, R395-R396.
2 PhD Research Studentships
Personalised Summary Generation
http://www.jobs.ac.uk/job/ACE550/phd-research-studentship-personalised-summary-generation/
Machine Learning Methods for User Modelling and Personalised Summarisation
http://www.jobs.ac.uk/job/ACE552/phd-research-studentship-machine-learning-methods-for-user-modelling-and-personalised-summarisation/
Natural Language Processing Research Group, University of Sheffield – Department of Computer Science
Applications are invited for two fully funded PhD studentships on the topics stated above.
Application closing date is 4 March 2011.
Further particulars on each of the studentships as well as eligibility criteria are available from the URLs above.
Candidates should have a First Class Honours or a good 2.1 degree in Computer Science or Mathematics and have excellent computer programming skills. Experience with natural language processing is essential, and detailed knowledge of machine learning, text summarisation and/or natural language generation would be highly desirable. Research experience with Facebook, Twitter, and other social media would also be desirable, but is not strictly necessary, as would be knowledge of GATE.
For further information please contact Dr Kalina Bontcheva (K.Bontcheva(at)dcs.shef.ac.uk) or Dr Trevor Cohn (tcohn(at)dcs.shef.ac.uk) (only for the studentship on machine learning).
Applicants should apply using the online application form at:
http://www.shef.ac.uk/postgraduate/research/apply
Call For Papers: Social Web Mining, co-located with IJCAI, Barcelona, Spain, 18 July 2011
WWW: http://users.cecs.anu.edu.au/~sguo/swm.html
Introduction:
There is increasing interest in social web mining, as we can see from the ACM workshop on Social Web Search and Analysis. It is not until recently that great progresses have been made in mining social network for various applications, e.g., making personalized recommendations. This workshop focuses on the study of diverse aspects of social networks with their applications in domains including mobile recommendations, service providers, electronic commerce, etc.
Social networks have actually played an important role in different domains for about a decade, particularly in recommender systems. In general, traditional collaborative filtering approaches can be considered as making personalized recommendations based on implicit social interaction, where social connections are defined by some similarity metrics on common rated items, e.g., movies for the Netflix Prize.
With the recent development of Web 2.0, there emerges a number of globally deployed applications for explicit social interactions, such as Facebook, Flickr, LinkedIn, Twitter, etc. These applications have been exploited by academic institutions and industries to build modern recommender systems based on social networks, e.g., Microsoft’s Project Emporia that recommends tweets to user based on their behaviors.
In recent years, rapid progress has been made in the study of social networks for diverse applications. For instance, researchers have proposed various tensor factorization techniques to analyze user-item-tag data in Flickr for group recommendations. Also, researchers study Facebook to infer users’ preferences.
However, there exist many challenges in mining social web and its application in recommender systems. Some are:
* What is the topology of social networks for some specific application like LinkedIn?
* How could one build optimal models for social networks such as Facebook?
* How can one handle the privacy issue caused by utilizing social interactions for making recommendation?
* How could one model a user’s preferences based on his/her social interactions?
We hope to gather scientific researchers and industry in order to discuss the challenges, exchange ideas, and promote collaborations across different groups.
Topics:
The workshop will seek submissions that cover social networks, data mining, machine learning, and recommender systems. The workshop is especially interested in papers that focus on applied domains such as web mining, mobile recommender systems, social recommender systems, and privacy in social web mining. The following list provides examples of the types of areas in which we encourage submissions. The following comprises a sample, but not complete, listing of topics:
* Active learning
* Matchmaking
* Mobile recommender systems
* Multi-task learning
* Learning graph matching
* Learning to rank
* Online and contextual advertising
* Online learning
* Privacy in social networks
* Preference learning or elicitation
* Social network mining
* Social summarization
* Tag recommendation
* Transfer learning
* Web graph analysis
Program:
The workshop program consists of four invited talks, a number of oral presentations, a poster session, and a panel discussion session. Detailed information is given as follows:
* Invited Speakers
o Ricardo Baeza-Yates , Yahoo! Research Barcelona, Spain
o Bhaskar Mehta , Google Zurich, Switzerland
o Jurgen Van Gael, Microsoft Research Cambridge, UK
o Qiang Yang, Hong Kong University of Science and Technology, China
* Oral and Poster Presentation: details coming soon
Submission and Key Dates:
We use the EasyChair for paper submission. The paper format is the same as that of IJCAI 2011 ( Click here for details ), and the maximum number of pages is 10. Please note that a selection of workshop papers will be invited to a special issue of ACM Transactions on Intelligent Systems and Technology.
* Submission deadline: 5 April 2011
* Author notification: 5 May 2011
* Camera ready: 25 May 2011
* Workshop date: 18 July 2011
Program Co-Chair:
* Francesco Bonchi, Yahoo! Research Barcelona, Spain
* Wray Buntine, NICTA – ANU, Australia
* Ricard Gavalda, Technical University of Catalonia, Spain
* Shengbo Guo, Xerox Research Centre Europe, France
Program Committee:
* Tiberio Caetano, NICTA – ANU, Australia
* Wei Chen, Microsoft Research Asia, China
* Peter Christen, Australian National University, Australia
* Nello Cristianini, University Of Bristol, UK
* Hakim Hacid, Alcatel-Lucent Bell Labs, France
* Jian Huang, Google Pittsburgh, USA
* Jure Leskovec, Stanford University, USA
* Ernesto William De Luca, Technical University of Berlin – DAI-Labor, Germany
* Sherif Sakr, NICTA – UNSW, Australia
* Scott Sanner, NICTA – ANU, Australia
* Fabrizio Silvestri, ISTI CNR, Italy
* Julia Stoyanovich, University of Pennsylvania, USA
* Aixin Sun, National University of Singapore, Singapore
* Antti Ukkonen, Yahoo! Research Barcelona, Spain
* Jie (Jessie) Yin, CSIRO, Australia
* Yi Zhang, University of California, Santa Cruz, USA
Workshop Contact:
Email: Shengbo (dot) Guo@xrce.xerox.com
+33 (0)4 76 61 50 47 ; ; (Phone)
+33 (0)4 76 61 50 99 (Fax)
Mailing address: 6, chemin de Maupertuis, 38240 Meylan, France
For further information, please visit our website:
http://users.cecs.anu.edu.au/~sguo/swm.html
Workshop on Machine Listening in Multisource Environments (CHiME 2011)
In conjunction with Interspeech 2011, September 1st, 2011, Florence, Italy
http://www.dcs.shef.ac.uk/spandh/chime/workshop
————————————————-
Important Dates:
* Deadline for submission of papers: April 14th, 2011
* Notification of acceptance: June 2nd, 2011
* Final version of submission: June 14th, 2011
* Workshop: September 1st, 2011
Overview:
CHiME 2011 will consider the challenge of developing machine listening applications for operation in multisource environments, i.e. real-world conditions with acoustic clutter, where the number and nature of the sound sources is unknown and changing over time. CHiME will bring together researchers from a broad range of disciplines (computational hearing, blind source separation, speech recognition, machine learning) to discuss novel and established approaches to this problem. The cross-fertilisation of ideas will foster fresh approaches that efficiently combine the complementary strengths of each research field.
The workshop will also be hosting the PASCAL CHiME Speech Separation and Recognition Challenge. For more information please visit the Challenge Website (http://www.dcs.shef.ac.uk/spandh/chime/challenge).
Call for Participation:
We invite original submissions for oral or poster presentation during the workshop. Relevant research topics include (but are not limited to),
* automatic speech and music processing in multisource environments,
* acoustic event detection in multisource environments,
* sound source detection and tracking in multisource environments,
* music information retrieval in multisource environments,
* sound source separation or enhancement in multisource environments,
* robust feature extraction and classification in multisource environments,
* scene analysis and understanding for multisource environments.
Abstracts or full papers are to be submitted by 14th April. After the workshop participants will be invited to submit extended versions of their papers to a peer-reviewed *special issue* of the journal, Computer Speech and Language on the theme of Multisource Environments.
Organising Committee:
Dr Jon Barker, University of Sheffield, UK
Dr Emmanuel Vincent, INRIA Rennes, France
Prof. Dan Ellis, Columbia University, USA
Prof. Phil Green, University of Sheffield, UK
Dr. John Hershey, Mitsubishi Electric Research Laboratories, USA
Prof. Walter Kellermann, University of Erlangen-Nuremberg, Germany
Prof. Hiroshi Okuno, Kyoto University, Japan
—
