[Help]  >>
User Login | Event Admin Login
Category: List of Bases 2012 2012 ICTP Activities in Trieste ICTP Activities in Trieste
Session Menu
lecture

Cosponsor(s):

SNN Adaptive Intelligence, Radboud University of Nijmegen, The Netherlands

Start Time:

20 August 2012 at 08:30

Ends On:

31 August 2012

Location:

Trieste - Italy

Venue:

AGH (Kastler Lecture Hall)

Organizer(s):

H.J. Kappen, M. Opper, R. Zecchina. Local Organiser: M. Marsili

Description:

Machine learning is a branch of computer science that addresses the issue of automatically recognizing complex patterns in empirical data and make intelligent decision based on the representation of these patterns. Optimal solutions are the result of the interplay between reproducing empirical observations and providing answers that generalize optimally for yet unobserved data. This interplay is similar to that between energy and entropy in statistical mechanics. Many applications of machine learning deal with large problems with many variables, and empirical data (training set) is limited to a very small fraction of the space of possible inputs. In statistical mechanics jargon, the entropic contribution plays a dominant role.

Concepts and tools in statistical physics have been proved useful in applications to ensembles of problems in computer science, such as constraint satisfaction or error-correcting codes, by revealing the structure of the space of solutions in typical cases, assessing the performance of algorithms and suggesting new efficient heuristics. The School and Workshop aim at investigating the relevance of these concepts and techniques for the broader field of machine learning, exploring opportunities of mutual cross fertilization between the two communities.

The School provides an overview of the key concepts of Machine Learning which are central to large scale applications. Lectures will be delivered by experts in machine learning and target an audience of researchers and graduate students with a background in physics and/or mathematics. It will last for a week and a half, and be followed by a two days Workshop. Both the School and Workshop will have a particular emphasis on applications of machine learning to computational biology.

TOPICS COVERED BY THE SCHOOL:
• Graphical models
• Approximate inference
• Control theory and reinforcement learning
• Neural networks
• Non-Bayesian kernel methods
• Algorithms for Bayesian kernel machines on large scale problems
• Nonparametric Monte Carlo techniques and Dirichelet processes
• Machine learning in systems biology

Applicants willing to present a poster are invited to submit a short one-page abstract along with their online application form.

Activity Secretariat:
The Abdus Salam International Centre for Theoretical Physics (ICTP)
Strada Costiera 11 
I-34151 Trieste
Italy

Telefax: +39-040-224163
 Telephone: +39-040-2240305


smr2361@ictp.it

DEADLINE for receipt of applications: 20 APRIL 2012




Material:

20 August 2012
08:30
09:45
Registration to the Workshop - Adriatico Guest House, lower level 1
01h15'
09:45
10:00
Opening Remarks by Organizers
15'
10:00
11:00
Introduction to probabilistic and Bayesian inference - I
01h00'
M. OPPER
TU Berlin, Germany
11:00
11:30
Coffee break
30'
11:30
12:30
Kernel methods - I
01h00'
G. LEVER
UC London, U.K.
12:30
14:00
Lunch break
01h30'
14:00
15:00
Introduction to probabilistic and Bayesian inference - II
01h00'
M. OPPER
TU Berlin, Germany
15:00
16:00
Kernel methods - II
01h00'
G. LEVER
UC London, UK
19:30
21:00
WELCOME RECEPTION - Adriatico Terrace
01h30'
09:00
10:00
Introduction to probabilistic and Bayesian inference - III
01h00'
M. OPPER
TU Berlin, Germany
10:00
11:00
Bayesian nonparametrics - I
01h00'
Y.W. TEH
UC London, U.K.
11:00
11:30
Coffee break
30'
11:30
12:30
Bayesian nonparametrics - II
01h00'
Y.W. TEH
UC London, U.K.
12:30
14:00
Lunch break
01h30'
14:00
15:00
Kernel methods - III
01h00'
G. LEVER
UC London, U.K.
15:00
16:00
Introduction to probabilistic and Bayesian inference - IV
01h00'
M. OPPER
TU Berlin, Germany
09:00
10:00
Kernel methods - IV
01h00'
G. LEVER
UC London, U.K.
10:00
11:00
Bayesian nonparametrics - III
01h00'
Y.W. TEH
UC London, U.K.
11:00
11:30
Coffee break
30'
11:30
12:30
Variational methods, duality and message-passing - I
01h00'
M. WAINWRIGHT
UC Berkeley, U.S.A.
12:30
14:00
Lunch break
01h30'
09:00
10:00
Bayesian nonparametrics - IV
01h00'
Y.W. TEH
UC London, U.K.
10:00
11:00
Variational methods, duality and message-passing - II
01h00'
M. WAINWRIGHT
UC Berkeley, U.S.A.
11:00
11:30
Coffee break
30'
11:30
12:30
Machine Learning in systems biology - I
01h00'
O. WINTHER
DTU Lyngby & KU Copenhagen, Denmark
12:30
14:00
Lunch break
01h30'
14:00
15:00
Variational methods, duality and message-passing - III
01h00'
M. WAINWRIGHT
UC Berkeley, U.S.A.
15:00
16:00
Machine Learning in systems biology - II
01h00'
O. WINTHER
DTU Lyngby & KU Copenhagen, Denmark
09:00
10:00
Machine Learning in systems biology - III
01h00'
O. WINTHER
DTU Lyngby & KU Copenhagen, Denmark
10:00
11:00
Variational methods, duality and message-passing - IV
01h00'
M. WAINWRIGHT
UC Berkeley, U.S.A.
11:00
11:30
Coffee break
30'
11:30
12:30
Machine Learning in systems biology - IV
01h00'
O. WINTHER
DTU Lyngby & KU Copenhagen, Denmark
12:30
14:00
Lunch break
01h30'
14:00
14:05
Afternoon Free
05'
09:00
10:00
Stochastic optimal control theory and its link to statistical physics - I
01h00'
B. KAPPEN
Nijmegen, The Netherlands
10:00
11:00
Large scale variational Bayesian inference and decision making for continuous variable models - I
01h00'
M.W. SEEGER
EPFL, Lausanne, Switzerland
11:00
11:30
Coffee break
30'
11:30
12:30
Stochastic optimal control theory and its link to statistical physics - II
01h00'
B. KAPPEN
Nijmegen, The Netherlands
12:30
14:00
Lunch break
01h30'
14:00
15:00
Large scale variational Bayesian inference and decision making for continuous variable models - II
01h00'
M.W. SEEGER
EPFL, Lausanne, Switzerland
15:00
16:00
Stochastic optimal control theory and its link to statistical physics - III
01h00'
B. KAPPEN
Nijmegen, The Netherlands
09:00
10:00
Large scale variational Bayesian inference and decision making for continuous variable models - III
01h00'
M.W. SEEGER
EPFL, Lausanne, Switzerland
10:00
11:00
Stochastic optimal control theory and its link to statistical physics - IV
01h00'
B. KAPPEN
Nijmegen, The Netherlands
11:00
11:30
Coffee break
30'
11:30
12:30
Large scale variational Bayesian inference and decision making for continuous variable models - IV
01h00'
M.W. SEEGER
EPFL, Lausanne, Switzerland
12:30
14:00
Lunch break
01h30'
14:00
14:05
Afternoon Free
05'
29 August 2012
09:00
10:00
Registration to the Workshop for NEW COMERS
01h00'
10:00
10:10
Opening Remarks
10'
10:10
10:50
The challenges and opportunities in using high-throughput data to understand how cells work
40'
E. FRAENKEL
MIT Cambridge, U.S.A.
10:50
11:20
Coffee break
30'
11:20
12:00
On the first eigenvalue/eigenvector in sparse random symmetric matrices
40'
Y. KABASHIMA
Tokyo Institute of Technology, Yokohama, Japan
12:00
12:40
Spatially coupled codes
40'
R. URBANKE
EPFL, Lausanne, Switzerland
12:40
14:00
Lunch break
01h20'
14:00
14:40
GOING DATA-DRIVEN: From artificial intelligence to science and society
40'
N. CRISTIANINI
Bristol, U.K.
14:40
15:20
Inferring mesoscopic structure in complex networks with message passing
40'
J. REICHARDT
Univ. Wurzburg, Germany
15:20
15:50
Coffee break
30'
15:50
16:30
Convergent and scalable algorithms for expectation propagation approximate Bayesian inference
40'
M.W. SEEGER
EPFL, Lausanne, Switzerland
09:30
10:10
Network inference for directed graphs with stochastic hidden units
40'
J. HERTZ
Nordita, Copenhagen, Denmark
10:10
10:50
Network reconstruction with hidden nodes
40'
Y. ROUDI
NTNU, Trondheim, Norway
10:50
11:20
Coffee break
30'
11:20
12:00
Some recent results on the inverse Ising problem
40'
F. RICCI-TERSENGHI
La Sapienza, Roma, Italy
12:00
12:40
Polymer physics for route optimisation on the London underground
40'
D. SAAD
Aston Univ., Birmingham, U.K.
12:40
14:00
Lunch break
01h20'
14:00
14:40
Variational methods in inference and optimisation
40'
D. BARBER
UC London, U.K.
14:40
15:20
Phenomenological study on spatial coupling
40'
T. TANAKA
Graduate School of Informations, Kyoto University, Japan
16:30
17:30
MAIN LECTURE HALL - LEONDARDO BUILDING COLLOQUIUM "Occam's Razar in Massive Data Acquisition: A statistical Physics Approach"
01h00'
M. MEZARD
LPTMS, Univ. Paris Sud, France
09:30
10:10
The Variational Garrote
40'
B. KAPPEN
Nijmegen, The Netherlands
10:10
10:50
Perturbative corrections to expectation propagation
40'
M. OPPER
TU Berlin, Germany
10:50
11:20
Coffee break
30'
11:20
12:00
Exchangeable random partitions, Bayesian inference and species sampling problems
40'
A. LIJOI
Pavia, Italy
12:00
12:40
Pseudo-likelihood maximization for the protein reconstruction problem
40'
E. AURELL
KTH, Stockholm, Sweden
12:40
13:00
Closing Remarks
20'
If you want to make a direct link from your Web page to this agenda, please use this URL:
http://cdsagenda5.ictp.trieste.it/full_display.php?ida=a11184

Maintained by: The CDS Support Team (Bugs and reports)
This page is loaded in 0.35808300971985 seconds.