"Gal Elidan"

Gal Elidan
I am a senior lecturer in the Department of Statistics at the Hebrew University

Phone:
02-588-3313

e-mail:
galel at huji dot ac dot il

Address:
Department of Statistics
Faculty of Social Sciences, Room 4419
Mount Scopus Campus
The Hebrew University of Jerusalem

Research Interests:

Probabilistic Graphical Models. Fundamental representations and methods for inference and learning in large scale domains, with an emphasis on high-level elements such as the discovery of hidden variables and classes, transfer of knowledge between related classes/tasks, and the combination of discriminative and generative elements within a unified model.
Real-life Applications. Applying fundamental techniques to challenging domains such as computational biology and machine vision. Recently, I have started looking into the development of principled techniques based on probabilistic knowledge for diagnosis in the field of medical informatics. Using challenges in these fields as a feedback trigger for fundamental research.

Journal Publications

Shape-Based Object Localization for Descriptive Classification
with Geremy Heitz, Ben Packer, and Daphne Koller.
International Journal of Computer Vision, 2009.
Learning Bounded Treewidth Bayesian Networks (pdf).
with Stephen Gould,
Journal of Machine Learning Research, 2008.
Multi-Class Segmentation with Relative Location Prior (pdf).
with Stephen Gould, Jim Rodgers, David Cohen and Daphne Koller.
International Journal of Computer Vision, 2008.
Markov random field based automatic image alignment for electron tomography (pdf).
with Fernando Amat, Farshid Moussavi, Louis Comolli, Kenneth Downing and Mark Horowits.
Journal of Structural Biology, 2007.
"Ideal Parent" Structure Learning for Continuous Variable Networks (pdf).
with Iftach Nachman and Nir Friedman.
Journal of Machine Learning Research (JMLR), Vol. 8, p. 1799-1833, 2007.
Max-margin classification of incomplete data
with Gal Chechik, Geremy Heitz, Pieter Abeel, and Daphne Koller
Journal of Machine Learning Research (JMLR), In Press, 2007.
Towards an Integrated Protein-Protein Interaction Network: A Relational Markov Network Approach (pdf).
with Ariel Jaimovich, Hanah Margalit and Nir Friedman
Journal of Computational Biology (JCB), Mar 2006, Vol. 13, No. 2: 145-164.
Learning Hidden Variable Networks: The Information Bottleneck Approach (pdf)
with Nir Friedman.
Journal of Machine Learning Research (JMLR), Vol. 6, p. 81-127, 2005.
CIS: Compound Importance Sampling Method for Transcription Factor Binding Site p-value Estimation (pdf)
with Yoseph Barash, Nir Friedman and Tommy Kaplan.
BioInformatics, Vol. 20(6), p. 839-46, 2004.
Inferring Subnetworks from Preturbed Expression Profiles (pdf)
with Dana Pe'er, Aviv Regev and Nir Friedman
Bioinformatics, Vol. 17:S, p. 215-224, 2001.

Peer-Reviewed Conference Publications

Learning Bounded Treewidth Bayesian Networks
with Stephen Gould,
Neural Information Processing Systems (NIPS) conference, 2009.
Shape-Based Object Localization for Descriptive Classification (pdf)
with Ben Packer, Geremy Heitz, and Daphne Koller
Neural Information Processing Systems (NIPS) conference, 2009.
Convex Point Estimation using Undirected Bayesian Transfer Hierarchies (pdf)
with Ben Packer, Geremy Heitz, and Daphne Koller
The Twenty Fourth Conference on Uncertainty in Artificial Intelligence (UAI), 2008.
Using Combinatorial Optimization within Max-Product Belief Propagation (pdf)
with John Duchi, Danny Tarlow, and Daphne Koller
Neural Information Processing Systems (NIPS) conference, 2006.
Max-margin classification of incomplete data (pdf)
with Gal Chechik, Geremy Heitz, Pieter Abeel, and Daphne Koller
Neural Information Processing Systems (NIPS) conference, 2006.
Residual Belief Propagation: Informed Scheduling for Asynchronous Message Passing (pdf)
with Ian McGraw and Daphne Koller
Twenty Second Conference on Uncertainty in Artificial Intelligence (UAI), 2006.
Learning Object Shape: From Drawings to Images (pdf)
with Geremy Heitz and Daphne Koller
IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR), 2006.
Toward an Integrated Protein-Protein Interaction Network (pdf)
with Ariel Jaimovich and Nir Friedman.
The Ninth Annual International Conference on Research in Computational Molecular Biology (RECOMB), 2005 (also presented in the KEYSTONE 2005 workshop).
"Ideal Parent" Structure Learning for Continuous Variable Networks (pdf)
with Iftach Nachman and Nir Friedman.
Twentieth Conference on Uncertainty in Artificial Intelligence (UAI), 2004. Runner-up for Best Student paper award.
CIS: Compound Importance Sampling Method for Transcription Factor Binding Site p-value Estimation (pdf)
with Yoseph Barash, Nir Friedman and Tommy Kaplan.
International Conference on Intelligent Systems for Molecular Biology (ISMB), 2004.
The Information Bottleneck Expectation Maximization Algorithm (pdf)
with Nir Friedman,
Nineteenth Conference on Uncertainty in Artificial Intelligence (UAI), 2003.
Modeling Dependencies in Protein-DNA Binding Sites (pdf) (web supplement)
with Yoseph Barash, Nir Friedman and Tommy Kaplan.
The Seventh Annual International Conference on Research in Computational Molecular Biology (RECOMB), 2003.
Data Perturbation for Escaping Local Maxima in Learning (pdf) (poster)
with Matan Ninio, Nir Friedman and Dale Schuurmans.
The Eighteenth National Conference on Artificial Intelligence, 2002.
Inferring Subnetworks from Perturbed Expression Profiles (pdf)
with Dana Pe'er, Aviv Regev and Nir Friedman.
International Conference on Intelligent Systems for Molecular Biology (ISMB), 2001. Best paper award.
Learning the Dimensionality of Hidden Variables (pdf) (poster)
with Nir Friedman.
The Seventeenth Conference on Uncertainty in Artificial Intelligence (UAI), 2001.
Discovering hidden variables: A structure Based-Approach (pdf) (poster)
with Noam Lotner, Nir Friedman and Daphne Koller.
Neural Information Processing Systems (NIPS), 2000.

Dissertation

Ph.D. dissertation, Hebrew University, 2004

Teaching

Spring 2009: Bayesian Data Analysis
Winter 2008/9: BASICS: Probablity, Data and Computers
Fall 2005: Co-instructor with Daphne Koller of Probabilistic Method in AI
Winter 2003/04: TA (in class teaching) of Introduction to Probabilistic Graphical Models by Yair Weiss.
Winter 2002/03: TA (in class teaching) of Introduction to Probabilistic Graphical Models by Yair Weiss.
Winter 2001/02: TA (in class teaching) of Probabilistic Methods in AI by Nir Friedman.
Winter 2000/01: TA (in class teaching) of Probabilistic Methods in AI. by Nir Friedman.
Spring 2000: TA (in class teaching) of Data Structures