Reading Group

The reading group will convene every other week, on Mondays at 1:15PM. Meetings last one hour. Every week, a member of the group will present an important paper from the literature. Presentations can be very informal, and serve mainly as a starting point for discussions. People attending the meetings are not expected to read the paper beforehand, but to participate in the discussions. If you would like to present a paper, or have a suggestion to add a paper, please contact us at dubhashi at cs dot chalmers dot se or henkw at chalmers dot se. Schedule (newest on top)

1. May 30 2011, 1:15-PM-2:15PM
   1. Location: Blue 6349 (S2 Dept. 6th floor)
   2. Presenter: Devdatt Dubhashi and Vinay Jethava
   3. Topic: Inferring time dynamics in networks
   4. Papers:
      1. Q. Ho, L. Song and E. P. Xing, Evolving Cluster Mixed-Membership Blockmodel for Time-Evolving Networks, Proceedings of the 14th International Conference on Artifical Intelligence and Statistics (AISTAT 2011).
      2. M. Kolar and E. P. Xing, On Time Varying Undirected Graphs, Proceedings of the 14th International Conference on Artifical Intelligence and Statistics (AISTAT 2011).
      3. S. Hanneke, W. Fu and E. P. Xing, Discrete Temporal Models of Social Networks, Electronic Journal of Statistics Vol. 4 (2010) 585-605. (arXiv:0908.1258, communicated August 2009.)
      4. M. Kolar, L. Song, A. Ahmed, and E. P. Xing, Estimating Time-Varying Networks , Annals of Applied Statistics, Vol. 4, No. 1, 94 - 123, 2010. (earlier version appeared in arXiv:0812.5087)
      5. M. Kolar, L. Song and E. P. Xing, Sparsistent Learning of Varying-coefficient Models with Structural Changes, Proceeding of the 23rd Neural Information Processing Systems, (NIPS 2009). [appendix available soon]
2. May 3 2011, 1:15-PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept. 6th floor)
   2. Presenter: Ashkan Panahi
   3. Topic: LASSO-based estimation method and implementations
   4. Related papers:
      1. Banerjee, O., Ghaoui, L. E. & d'Aspremont, A. (2008), `Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data', Journal of Machine Learning Research 9, 485-516.
      2. El Ghaoui, L., Viallon, V. & Rabbani, T. (2010), Safe feature elimination in sparse supervised learning, Technical Report UC/EECS-2010-126, EECS Dept., University of California at Berkeley.
3. April 4 2011, 1:15PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept. 6th floor)
   2. Presenter: Lennart Svensson
   3. Topic: Bayesian sparse signal reconstruction [slides]
   4. Related paper: The Horseshoe Estimator for Sparse Signals Carlos M. Carvalho, Nicholas G. Polson, and James G. Scott University of Chicago GSB and Duke University
4. March 21 2011, 1:15PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept. 6th floor)
   2. Presenter: Henk Wymeersch
   3. Topic: Dykstra's Algorithm
   4. Related paper: An Iterative Procedure for Obtaining I-Projections onto the Intersection of Convex Sets, by Richard L. Dykstra, Source: Ann. Probab. Volume 13, Number 3 (1985), 975-984.
5. February 28 2011, 1:15PM-2:15PM
   1. Location: Landahlsrummet 3340 (S2 Dept, 3rd floor)
   2. Presenter: Giuseppe Durisi (S2)
   3. Topic: Compressed Sensing
   4. Related papers:
      1. Graphical Models in compressed sensing, A. Montanari 2010.
      2. http://arxiv.org/abs/1011.3854, "A probabilistic and RIPless theory of compressed sensing," Emmanuel J. Candes, Yaniv Plan
6. February 14 2011, 1:15PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept. 6th floor)
   2. Presenter: Vinay Jethava
   3. Topic: Continuous-time belief propagation
   4. Related papers:
      1. Continuous-time belief propagation, T. El-Hay, I. Cohn, N. Friedman, and R. Kupferman, Proc. Twenty Seventh International Conf. on Machine Learning (ICML 1
         1. 0), 2010.
      2. Continuous-time Bayesian Networks, Nodelman, Stanford PhD dissertation, June 2007.
      3. U. Nodelman, C.R. Shelton, and D. Koller (2002). "Continuous Time Bayesian Networks." Proceedings of the Eighteenth Conference on Uncertainty in Artificial Intelligence (UAI) (pp. 378-387).
      4. S. Saria, U. Nodelman, and D. Koller (2007). "Reasoning at the Right Time Granularity." Proceedings of the Twenty-third Conference on Uncertainty in AI (UAI).
7. January 31 2011, 1:15PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept, 6th floor)
   2. Presenter: Devdatt Dubhashi (CSE)
   3. Topic: Sparsity in Learning
   4. Related papers:
      1. Sparse Modeling, ICML tutorial, I. Rish, G. Grabarnik, 2010
      2. High-dimensional Ising model selection using ℓ1-regularized logistic regression, Pradeep Ravikumar, Martin J. Wainwright, and John D. Lafferty, Source: Ann. Statist. Volume 38, Number 3 (2010), 1287-1319.
      3. A unified framework for high-dimensional analysis of M-estimators with decomposable regularizers, Authors: Sahand Negahban, , Pradeep Ravikumar, Martin J. Wainwright, Bin Yu
8. January 17 2011, 1:15PM-2:15PM
   1. Location: Blue room 6349 (S2 Dept, 6th floor)
   2. Presenter: Henk Wymeersch (S2)
   3. Topic: Exact Inference as a intractable convex optimization problem
   4. Related paper: A new class of upper bounds on the log partition function, Wainwright, M.J.; Jaakkola, T.S.; Willsky, A.S.; TIT, 2005.

List of Potential Papers

We welcome suggestions for additional papers.

1. Factor graphs and the sum-product algorithm, Kschischang, F.R. ; Frey, B.J. ; Loeliger, H.-A. ; TIT 2001.
2. Constructing free-energy approximations and generalized belief propagation algorithms, Yedidia, J.S.; Freeman, W.T.; Weiss, Y.; TIT, 2005
3. Graphical Models, Exponential Families, and Variational Inference, MJ Wainwright, MI Jordan - Machine Learning, 2008
4. A new class of upper bounds on the log partition function, Wainwright, M.J.; Jaakkola, T.S.; Willsky, A.S.; TIT, 2005
5. Convergent Tree-Reweighted Message Passing for Energy Minimization, Kolmogorov, V., Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2006
6. Convergent message passing algorithms - a unifying view, Talya Meltzer, Amir Globerson and Yair Weiss, UAI 2009
7. A Message-Passing Paradigm for Optimization, C. C. Moallemi. Ph.D. thesis, Stanford University, 2007
8. Belief Propagation, Dykstra's Algorithm, and Iterated Information Projections, Walsh, J.M. ; Regalia, P.A. ;TIT 2010
9. A Direct Formulation of Sparse PCA using Semidefinite Programming. A d’Aspremont, L. El Ghaoui, M. Jordan, and G. Lanckriet. SIAM Review, vol. 49, no. 3, 2007.
10. High-dimensional Ising model selection using l_1 - regularized logistic regression, P. Ravikumar, Martin Wainwright, John Lafferty, Annals of Statistics, vol. 38, no. 3, 2010
11. Sparse inverse covariance estimation with the graphical lasso, as J. Friedman, T. Hastie, R. Tibshirani, Biostatistics 2008.
12. A spectral algorithm for learning Hidden Markov Models. D. Hsu, S. Kakade, T. Zhang, COLT 2009
13. Hilbert space embeddings and metrics for probability measures B. Sriperumbudur et al. JMLR, vol. 11, 2010
14. Hilbert space embeddings of Hidden Markov Models, Le Song et al. ICML 2010
15. Non-parametric density estimation: toward computational tractability, Gray and Moore, SIAM Intl. Conf. on Data Mining, 2003.
16. Convex Optimization with Sparsity-Inducing Norms, F. Bach, R. Jenatton, J. Mairal and G. Obozinski. In S. Sra, S. Nowozin, S. J. Wright., editors, Optimization for Machine Learning, MIT Press, 2011.
17. Convex Analysis and Optimization with Submodular Functions: a Tutorial, F. Bach. Technical report, HAL 00527714, 2010.
18. High-dimensional statistics: Some progress and challenges ahead, M. Wainwright; 2010
19. Graphical Models in compressed sensing, A. Montanari 2010.

Prerequisites

To ensure that we can have meaningful discussions, participants are expected to have some background in at least one of the topics of the reading group. For machine learning, inference, and graphical models, the following resources are a good start:

• Kevin Murphy's tutorial: http://www.cs.ubc.ca/~murphyk/Bayes/bnintro.html
• Zubin's video lectures: http://videolectures.net/mlss09uk_ghahramani_gm/
• Barber's book: http://www.cs.ucl.ac.uk/staff/d.barber/brmlhttp://www.cs.ucl.ac.uk/staff/d.barber/brml