e-LICO multi-omics prediction challenge with background knowledge on Obstructive Nephropathy

We present a biological data-mining problem that poses a number of significant challenges; the available data: (i) are of high dimensionality but of extremely small sample size, (ii) come from different sources which correspond to different biological levels, (iii) exhibit a high degree of feature dependencies and interactions within and between the different sources; some of the interactions between the different sources are known and available as background knowledge, and (iv) are incomplete.

This data was obtained from patients with Obstructive Nephropathy (ON) which is the most frequent nephropathy observed among newborns and children, and the first cause of end stage renal disease usually treated by dialysis or transplantation. The goal is to construct diagnostic models that accurately connect the biological levels to the severity of the pathology. We particularly welcome data mining approaches and learning methods that are able to accommodate the available background information in order to address the formidable challenge of high dimensionality small sample size of our setting and deliver better models.

A prize is envisaged for the top performing approaches (2500EU in total). The price is sponsored by Rapid-I the company that supports RapidMiner, probably the most popular open-source data mining environment, and the European Commission through the e-Lico EU project. Participants are expected to prepare a paper, maximum 8 pages, describing their approach. We plan to have a number of selected papers considered for publication in a special issue of a journal (to be announced soon).

Challenge web page: http://tunedit.org/challenge/ON .

Started: Sep 15, 2010
Ends: Dec 19, 2010

Organizing Committee:
– Alexandros Kalousis, University of Geneva, Switzerland
– Julie Klein, Inserm U858, Toulouse, France
– Joost Schanstra, Inserm U858, Toulouse, France
– Adam Woznica, University of Geneva, Switzerland

NIPS 2010 workshop: Monte Carlo Methods for Bayesian Inference in Modern Day Applications

Call for Contributions

—————————————————————————
NIPS 2010 workshop
Monte Carlo Methods for Bayesian Inference in Modern Day Applications
http://montecarlo.wikidot.com/
http://nips.cc/

December 10 or 11, 2010
Whistler, Canada. Westin Resort and Spa and Hilton Resort and Spa
—————————————————————————

We invite submissions on Monte Carlo methods and their practical
application. Particularly welcome are “tricks of the trade” and “war
stories” that might not make it into conventional publications.
Submissions are solicited both from researchers developing new
methodology and from practitioners using established techniques.

Send poster abstracts of up to one page to
montecarlo-2010@cs.toronto.edu
by Oct 31, 2010. Use the NIPS style file with no anonymity. We will
notify acceptances by Nov 5, and before the NIPS early registration
deadline.

We intend to invite key contributions from the workshop to submit full
papers to a JMLR W&CP issue to appear in the new year.

We also invite contributions to the wiki, including suggested readings
and discussion topics:
http://montecarlo.wikidot.com/

The organizers:
Ryan Prescott Adams, http://www.cs.toronto.edu/~rpa/
Mark Girolami, http://www.dcs.gla.ac.uk/inference/
Iain Murray, http://homepages.inf.ed.ac.uk/imurray2/

Confirmed invited speakers:
Derek Bingham
Arnaud Doucet
Andrew McCallum
Yee-Whye Teh

Workshop description:

Monte Carlo methods have been the dominant form of approximate inference for
Bayesian statistics over the last couple of decades. Monte Carlo methods are
interesting as a technical topic of research in themselves, as well as enjoying
widespread practical use. In a diverse number of application areas Monte Carlo
methods have enabled Bayesian inference over classes of statistical models which
previously would have been infeasible. Despite this broad and sustained
attention, it is often still far from clear how best to set up a Monte Carlo
method for a given problem, how to diagnose if it is working well, and how to
improve under-performing methods. The impact of these issues is even more
pronounced with new emerging applications. This workshop is aimed equally at
practitioners and core Monte Carlo researchers. For practitioners we hope to
identify what properties of applications are important for selecting, running
and checking a Monte Carlo algorithm. Monte Carlo methods are applied to a broad
variety of problems. The workshop aims to identify and explore what properties
of these disparate areas are important to think about when applying Monte Carlo
methods.

We look forward to seeing you in Whistler this December!

Research Fellow in Computer Vision and Machine Learning

Research Fellow in Computer Vision and Machine Learning
School of Computing, University of Leeds, UK

(Full Time, Fixed term 14 months)

You will work with Dr Mark Everingham (http://www.comp.leeds.ac.uk/me/) on an
EPSRC funded project investigating new methods for learning articulated human
pose estimation from weak or approximate supervision. The project has three
main aims: (i) developing machine learning methods to learn from approximate
annotation and “side information” for example simple models of human anatomy;
(ii) developing strong models of appearance to give robust pose estimation,
using the developed machine learning approach. This will include higher order
cues modelling appearance of limbs, dependencies between limbs and appearance
of joints and configurations of limbs; (iii) producing a large dataset of
approximately annotated consumer images, at least two orders of magnitude
larger than available datasets. Further information about the project can be
found online:

http://www.comp.leeds.ac.uk/me/Pose/

Applicants are expected to have a PhD (or to be awarded shortly) in a related
topic. You should have experience in developing and applying computer vision
and machine learning algorithms, especially probabilistic methods. Expertise
in graphical models, structured learning or human pose estimation would be a
particular advantage. You should be a proficient programmer in MATLAB and
C/C++. You should be self-motivated, good at time management and planning,
and have a proven ability to meet deadlines. Good communication and
presentation skills are also important.

Salary: Grade 7 (£29,853 – £35,646 p.a)

Job description and person specification:
http://hr.leeds.ac.uk/jobs/FileDownload.aspx?filename=312501%20-%20RF%20-%20Computer%20Vision.doc

Apply using: Application form, CV and Equal Opportunities Monitoring form

Download an application form:
http://www.leeds.ac.uk/hr/forms/recruitment/app_form.pdf
http://www.leeds.ac.uk/hr/forms/recruitment/newapplicationform.doc

Informal enquiries:

Dr Mark Everingham, email: m.everingham(at)leeds.ac.uk

Send completed applications to:
Judi Drew, email j.a.drew(at)leeds.ac.uk or by post to:

Judi Drew
School of Computing
University of Leeds
Leeds
LS2 9JT

Closing date: 22 September 2010 at 17:00 GMT

Anticipated interview date: 01 October 2010

MLSB 2010 – call for posters/registration

******************* Call for Posters/Registration **********************

MLSB 2010

The Fourth International Workshop on Machine Learning in Systems Biology

15-16 October 2010, Edinburgh, Scotland

***********************************************************************

http://mlsb10.ijs.si/

(apologies for multiple postings)

REGISTRATION IS NOW OPEN: http://tinyurl.com/2cp7xwc

MOTIVATION

Molecular biology and all the biomedical sciences are undergoing a
true revolution as a result of the emergence and growing impact of a
series of new disciplines/tools sharing the “-omics” suffix in their
name. These include in particular genomics, transcriptomics,
proteomics and metabolomics, devoted respectively to the examination
of the entire systems of genes, transcripts, proteins and metabolites
present in a given cell or tissue type.

The availability of these new, highly effective tools for biological
exploration is dramatically changing the way one performs research in
at least two respects. First, the amount of available experimental
data is not a limiting factor any more; on the contrary, there is a
plethora of it. Given the research question, the challenge has
shifted towards identifying the relevant pieces of information and
making sense out of it (a “data mining” issue). Second, rather
than focus on components in isolation, we can now try to understand
how biological systems behave as a result of the integration and
interaction between the individual components that one can now monitor
simultaneously (so called “systems biology”).

Taking advantage of this wealth of “genomic” information has become a
conditio sine qua non for whoever ambitions to remain competitive in
molecular biology and in the biomedical sciences in general. Machine
learning naturally appears as one of the main drivers of progress in
this context, where most of the targets of interest deal with complex
structured objects: sequences, 2D and 3D structures or interaction
networks. At the same time bioinformatics and systems biology have
already induced significant new developments of general interest in
machine learning, for example in the context of learning with
structured data, graph inference, semi-supervised learning, system
identification, and novel combinations of optimization and learning
algorithms.

The Workshop is organized as “core – event” of Pattern Analysis,
Statistical Modelling and Computational Learning – Network of Excellence
2 (PASCAL 2, http://www.pascal-network.org/)

OBJECTIVE

The aim of this workshop is to contribute to the cross-fertilization
between the research in machine learning methods and their
applications to systems biology (i.e., complex biological and medical
questions) by bringing together method developers and
experimentalists. We encourage submissions bringing forward methods
for discovering complex structures (e.g. interaction networks,
molecule structures) and methods supporting genome-wide data analysis.

LOCATION AND CO-LOCATION

The workshop will take place 15-16 October 2010 at the Edinburgh
International Conference Centre and the Informatics Forum of the
University of Edinburgh. It will be part of the wokshop program of
ICSB 2010, The 11th International Conference on Systems Biology
(11-14 OCT 2010, http://www.icsb2010.org.uk/).

POSTER SUBMISSION INSTRUCTIONS

We invite you to submit an abstract of up to 4 pages (minimum 1 page)
describing new or recently published (2010) results, formatted
according to the Springer Lecture Notes in Computer Science
style. Each extended abstract must be submitted online via the Easychair
submission system: http://www.easychair.org/conferences/?conf=mlsb10

KEY DATES

30th September: Poster submission deadline.
REGISTRATION IS NOW OPEN: ttp://tinyurl.com/2cp7xwc

TOPICS

A non-exhaustive list of topics suitable for this workshop is given
below:

Methods

Machine learning algorithms
Bayesian methods
Data integration/fusion
Feature/subspace selection
Clustering
Biclustering/association rules
Kernel methods
Probabilistic inference
Structured output prediction
Systems identification
Graph inference, completion, smoothing
Semi-supervised learning

Applications

Sequence annotation
Gene expression and post-transcriptional regulation
Inference of gene regulation networks
Gene prediction and whole genome association studies
Metabolic pathway modeling
Signaling networks
Systems biology approaches to biomarker identification
Rational drug design methods
Metabolic reconstruction
Protein function and structure prediction
Protein-protein interaction networks
Synthetic biology

INVITED SPEAKERS (confirmed)

Florence d’Alche Buc, Universite d’Evry-Val d’Essonne, Evry, France
Nir Friedman, The Hebrew University of Jerusalem, Jerusalem, Israel
Ursula Kummer, BIOQUANT, University of Heidelberg, Germany
Hans Lehrach, Max Planck Institute for Molecular Genetics, Berlin, Germany
Vebjorn Ljosa, The Broad Institute of MIT and Harvard, USA

MLSB10 PROGRAM CHAIRS

Saöo Dûeroski, Jozef Stefan Institute, Ljubljana, Slovenia
Simon Rogers, University of Glasgow, UK
Guido Sanguinetti, University of Sheffield/University of Edinburgh, UK

SCIENTIFIC PROGRAM COMMITTEE (tentative)

Florence d’AlchÈ-Buc, University of Evry, France
Paolo Frasconi, Universit‡ degli Studi di Firenze, Italy
Cesare Furlanello, Fondazione Bruno Kessler, Trento, Italy
Pierre Geurts, University of LiËge, Belgium
Mark Girolami, University of Glasgow, UK
Dirk Husmeier, Biomathematics & Statistics Scotland, UK
Samuel Kaski, Helsinki University of Technology, Finland
Ross D. King, Aberystwyth University, UK
Neil Lawrence, University of Manchester, UK
Elena Marchiori, Vrije Universiteit Amsterdam, The Netherlands
Yves Moreau, Katholieke Universiteit Leuven, Belgium
William Stafford Noble, University of Washington, USA
Gunnar R‰tsch, FML, Max Planck Society, T¸bingen
Juho Rousu, University of Helsinki, Finland
CÈline Rouveirol, University of Paris XIII, France
Yvan Saeys, University of Gent, Belgium
Ljupco Todorovski, University of Ljubljana, Slovenia
Koji Tsuda, Max Planck Institute, Tuebingen
Jean-Philippe Vert, Ecole des Mines, France
Louis Wehenkel, University of LiËge, Belgium
Jean-Daniel Zucker, University of Paris XIII, France
Blaz Zupan, University of Ljubljana, Slovenia

LOCAL ORGANIZATION

Fiona Clark, University of Edinburgh, UK
Dragi Kocev, Jozef Stefan Institute, Ljubljana, Slovenia (webmaster)
—
Dr Simon Rogers
Lecturer in Inference
School of Computing Science
University of Glasgow
simon.rogers (at) glasgow.ac.uk
skype: sdrogersskype

Call for additional partners of PASCAL 2 Network of Excellence

PASCAL 2 is a European Commission financed FP7 Network of Excellence that started in March 2008 and that will run until February 2013. It focuses on the development of principled approaches to machine learning, statistical analysis and pattern recognition and their applications to diverse areas including cognitive systems and robotics, user interfaces, and vision, speech and text understanding. Many of the leading European research groups in these areas participate. The funding is allocated dynamically by a range of network programmes that promote network themes and encourage interaction and community building in these areas. As a part of this activity, we have held an annual call for new sites for the past two years, and we will hold a limited call again this year. Any site that is eligible to participate in EC financed projects can apply, but the process is highly selective and for budgetary and network-focus related reasons only a very small number of new sites can be admitted this year. New sites are expected to be world leading in their research areas and they must demonstrate links with existing sites that are sufficient to ensure future collaboration and good integration into the network. Sites working on any network-related theme can apply, but for this call preference will be given to building critical mass in our thematic priority area of creating artificial cognitive systems (including but not limited to robotic systems, embedded agents, etc) based on sound statistical and machine learning principles. For more information about the application process and deadlines, see below.

———————————————————————
Information required for an application to become a new PASCAL 2 site

For consideration at the 3 September 2010 Steering Committee meeting, applications including all of the following information must reach the Committee by FRIDAY 6 AUGUST 2010 and the corresponding letters of reference must reach the Committee by FRIDAY 13 AUGUST 2010. There will be no extensions. Applications and letters should be sent in PDF or plain text format to the PASCAL administrator Rebecca Martin at UCL, Rebecca.Martin at cs.ucl.ac.uk.

Information to include:

• The site name and manager.
• The name of the PASCAL 2 Beneficiary that will manage the site’s budget for it (for most sites this will be UCL [1])
• The names of two previously agreed sponsors from within the PASCAL 2 network. These must be from two different countries (one country can be that of the candidate site) and they must be permanent staff at their respective institutions. The applicant should ask each sponsor to email a confidential letter of support for the application to the Committee by FRIDAY 13 AUGUST 2010. The letters should evaluate the leadership and excellence of the site and the added value for the network as a whole of including it (complementarity, critical mass, probable collaborations, advancement of network goals…)
• A proposed initial list of members at the site, giving their employment status (permanent, postdoc, PhD student) and keywords for their research interests.
• A few paragraphs (absolute maximum 1 page) explaining concisely how the site intends to contribute to PASCAL 2 and describing the kind of work that it does, its alignment with PASCAL 2 and its goals, any past or current participations in PASCAL activities, how the group plans to interact with other PASCAL 2 sites,… Include a list of at most 5 recent publications that illustrate the quality and alignment of the work at the site. Small sites will need to argue that they have the critical mass to participate as an autonomous PASCAL site.
• The information required for a Contract Preparation Form: official names and addresses of the institution(s) concerned; names, email addresses and telephone numbers of the people in their contracts offices who will handle contract preparation; full names and titles of their contract signatories.
• Before submitting an application, you must also verify that your institution has read the Non-Beneficiary accession agreement and is willing to accede to it (Downloadable version available here).

———————————————————————-

[1] Most new sites are expected to join as “Non-Beneficiary” members. This essentially allows full network participation except that the site budget is managed by UCL rather than the site itself and staff can not be employed. If staff need to be employed later (e.g. following a successful Pump Priming or Harvest request) the site can apply for promotion to full Beneficiary status at that stage. However sites that are attached to existing multi-site Beneficiaries, notably the CNRS, can immediately become full members via their Beneficiary, subject to the agreement of that Beneficiary. The application and approval process is the same for all new sites.

Call for new members of PASCAL2 – update to contact details

It appears that the address for the applications to be sent to did not appear in the orginal advert.

Everything should be sent to Rebecca Martin – rebecca.martin at cs.ucl.ac.uk

The original advert has also been updated, and is available here: http://www.pascal-network.org/?q=node/164

Apologies for any inconvenience.

Postdoctoral position within the area of Distributed Text Mining at the Norwegian University of Science and Technology

A post-doctoral position in the area of distributed text mining within the
COMIDOR project is available at the Department of Computer and Information
Science (IDI) in the Faculty of Information Technology, Mathematics and
Electrical Engineering (IME), the Norwegian University of Science and
Technology (NTNU).

COMIDOR (Cooperative Mining of Independent Document Repositories) is a
research project at NTNU that will focus on the mechanisms and algorithms
necessary to perform mining of independent document repositories. COMIDOR is
funded by the Norwegian Research Council.

Traditionally, text mining has been performed on a single text
collection, and in the case of collections from several repositories
these collections have first been merged before performing the mining
process. In some application areas, merging of collections is not
acceptable. For example, some repositories can not be merged for legal
reasons, while some can not be merged because of risk of revealing
classified information. In the COMIDOR project the aim is to develop
new solutions to mining of independent document repositories without
communication of base repositories. A more detailed description of the
project can be found at http://research.idi.ntnu.no/comidor/.

The candidate should have a Ph.D. in Computer Science with solid
knowledge of text mining, data mining, or distributed data management, and
familiarity with program development. The postdoc position is for two
years.

For more information, please Prof. Kjetil Norvag (project leader), noervaag
at idi.ntnu.no.

Specific Conditions: The postdoc position is placed in Norwegian salary code
1352, gross NOK 446.700 per year (equivalent to approx. EUR 56.500 and USD
75.600).

Living Costs in Norway: A rule of thumb is that one pays appr. 35% tax on
net income. Public health care system is free for all (due to the tax
level). Generally there is no need for private health care insurances.
Education on all levels is free for all. Housing normally cost between NOK
4000 (for a simple one bed studio with kitchenette, possibly shared
bathroom) and 8000 (for a 3-4 piece apartment suitable for a small family)
per month.

Postdoctoral Position in Machine Learning with,application to Computer Vision at KTH, Stockholm, Sweden

A postdoctoral position in Machine Learning with application
to Computer Vision is available in the Computer Vision group
(CVAP) at KTH, Stockholm.

The position will be associated with an ongoing project in,
Computer Vision that can be either:

1. Wearable Visual Information Systems

http://www.nada.kth.se/cvap/vinst.html

This project aims at the processing of visual information captured in
wearable cameras. The main objectives are recognition, classification,
location finding in e.g. mobile phone cameras as well as the automatic
creation of visual diaries from continuously wearable cameras.

2. Capturing and Visualizing Large Scale Human Action

http://www.csc.kth.se/~sullivan/actvis/results.html

This project is also closely associated with the European project
FINE. http://www.projectfine.eu/

This project has as its main goal the mapping of human sports action
from video into 3D animations. Specifically we will consider the 3D
capture of players actions in football games. The project is carried
out in connection with the Swedish industrial partner TRACAB as well as
other industrial partners in Europe.

The length of the position is 2 years with possible extension up to 4
years. The applicant is expected to perform full time research of
which approximately 50% will be devoted to the associated project.

Applications should include a cover letter indicating the associated
research project of interest, a curriculum vitae, and the contact
information of references.

The position is part of a larger (3) post-doc program at KTH

Please follow the link:
http://www.kth.se/om/work-at-kth/vacancies/postdoctoral-positions-in-machine-learning-1.62352?l=en_UK

for detailed instructions of application

Deadline for application is October 1st 2010

For information contact:
Stefan Carlsson, stefanc at csc.kth.se, Tel: +46 8 790 8432
Josephine Sullivan sullivan at csc.kth,se +46 8 790 6136

Book Announcement: “Reinforcement Learning and Dynamic Programming Using Functions Approximators”

Dear machine learning researchers,

We are pleased to announce the recent release of our book:
“Reinforcement Learning and Dynamic Programming Using Functions Approximators”
(Lucian Busoniu, Robert Babuska, Bart De Schutter, and Damien Ernst)
in the Automation and Control Engineering series of Taylor & Francis CRC Press.

Book information:

Reinforcement learning (RL) can optimally solve decision and control problems involving complex dynamic systems, without requiring a mathematical model of the system. If a model is available, dynamic programming (DP), the model-based counterpart of RL, can be used. RL and DP are applicable in a variety of disciplines, including artificial intelligence, automatic control, economics, and medicine. Recent years have seen a surge of interest RL and DP using compact, approximate representations of the solution, which enable algorithms to address realistic problems.

This book provides an in-depth introduction to RL and DP with function approximators, with a focus on continuous-variable control problems. A concise description of classical RL and DP (Chapter 2) builds the foundation for the remainder of the book. This is followed by an extensive review of the state-of-the-art in RL and DP with approximation, which combines algorithm development with theoretical guarantees, illustrative numerical examples, and algorithm comparisons (Chapter 3). Each of the final three chapters (4 to 6) is dedicated to a representative algorithm from the authors’ research. These three algorithms respectively belong to the three major classes of methods: approximate value iteration, approximate policy iteration, and approximate policy search. The features and performance of these algorithms are highlighted in comprehensive experimental studies on a range of control applications.

Features:
* A concise introduction to the basics of RL and DP
* A detailed treatment of RL and DP with function approximators, including theoretical results and illustrative examples
* A thorough treatment of policy search techniques
* Comprehensive experimental studies on a range of control problems, including real-time control results
* An extensive, illustrative convergence and consistency analysis of an approximate value iteration algorithm

For graduate students and others new to the field, this book offers a thorough introduction to both the basics and emerging methods. And for those researchers and practitioners working in the fields of optimal and adaptive control, machine learning, artificial intelligence, and operations research, this resource offers a combination of practical algorithms, theoretical analysis, and comprehensive examples that they will be able to adapt and apply to their own work.

Access the book’s website at http://www.dcsc.tudelft.nl/rlbook/ for additional information, including computer code used in the experimental studies, information about ordering the book, etc.

Hoping you will find this book useful,

Sincerely,

The authors

Job opening: Research Fellow in Machine Learning and Corpus Classification

The Centre for Translation Studies, University of Leeds, UK, is inviting
applications for a Research Fellow post to work on a research project
funded by an FP7 grant from the European Commission. It is a full-time,
fixed term position to be filled as soon as possible, lasting until
31 Dec 2012.

We’re interested in a candidate with a PhD in Computational linguistics
or Machine Learning and experience in designing ML algorithms for text
categorisation tasks. For a formal application, take a look at the job
description:
http://hr.leeds.ac.uk/jobs/ViewJob.aspx?SId=16&JId=1306

I will be attending ACL2010 in Uppsala and SIGIR2010 in Geneva, so feel
free to contact me there informally.

Closing date for applications: 03 August 2010
Salary: Grade 7 (£29,853 – £35,646 p.a)
It is likely that an appointment will be made at or below £30,747 p.a.
since there are funding limitations which dictate the level at which the
appointment can start.