CS-295 Reasoning and Modeling with Graphical Models, Winter 2011

Classroom: DBH 1431 Day: Friday Time: 1:00 - 3:20p Instructor: Rina Dechter - dechter@ics.uci.edu The seminar will focus on recent advances in graphical models reasoning and knowledge representation, as well as on exploring some application areas. The main subject of seminar will be presenting paper from the recent NIPS conference and workshops Each student will be involved in presenting papers/chapters and will be encouraged to focus on a project of interest. Students will write a final report for their research project. Recommended papers: Vibhav Gogate, William Webb, Pedro Domingos. Learning Efficient Markov Networks. Abhay Jha, Vibhav Gogate, Alexandra Meliou, Dan Suciu. Lifted Inference Seen from the Other Side : The Tractable Features. Daniel Lowd, University of Oregon; Pedro Domingos. Approximate Inference by Compilation to Arithmetic Circuits. University of Washington. Martin Zinkevich, Markus Weimer, Alex Smola, and Lihong Li. Parallelized Stochastic gradient descent. Yahoo Research. Akshat Kumar & Shlomo Zilberstein. MAP Estimation for graphical Models by Likelihood Maximization. UMass Gal Elidan. Copula bayesian Networks. Hebrew University Jim C Huang, Nebojsa Jojic and Chris Meek. Exact inference and learning for cumulative distribution functionson loopy graphs. Microsoft Research Merixtell Vinyals, Jesus Cerquides and Juan Antonio Rodriguez-Aguilar,IIIA-CSIC; Alessandro Farinelli, University of Verona. Worst-case bounds on the quality of max-product fixed-points. M. Kumar, B. Packer, D. Koller. Self-Paced Learning for Latent Variable Models. S. Jakkam Reddi, S. Sarawagi, S. Vishwanathan. MAP estimation in Binary MRFs via Bipartite Multi-cuts. Corinna Cortes, Yishay Mansour, Mehryar Mohri. Learning Bounds for Importance Weighting. A. Bouchard-Cote, M. Jordan. Variational Inference over Combinatorial Spaces. Kamalika Chaudhuri, Sanjoy Dasgupta. Rates of convergence for the cluster tree. Workshops: Workshop on monte carlo methods. Workshop on learning on cores, clusters and clouds. Schedule

Week	Date	Topic	Readings
Week 1	1/7
Week 2	1/14	Andrew Gelfand: Terminating Orderings. Histograms of min-fill based tree-width William Lam: AOMDDs: PPTX , PDF
Week 3	1/21	William: AOMDDs: PPTX , PDF Rina
Week 4	1/28	Natasha: Akshat Kumar & Shlomo Zilberstein. MAP Estimation for graphical Models by Likelihood Maximization. UMass, Slides: PDF Lars: Anytime Depth-First Search versus Problem Decomposition. Submitted to IJCAI 2011.
Week 5	2/4	Lars: Jegelka, Bilmes. Online algorithms for submodular minimization with combinatorial constraints Teague: Reddi S.J., Sarawagi S., Vishwanathan S.; MAP estimation in Binary MRFs via Bipartite Multi-cuts; NIPS 2010. Slides
Week 6	2/11	Andrew: Joseph Gonzalez, Yucheng Low, and Carlos Guestrin (2009). "Residual Splash for Optimally Parallelizing Belief Propagation." In Artificial Intelligence and Statistics (AISTATS). Slides Rina Dechter: Finding the m best solutions using search. Slides
Week 7	2/18	Qiang: Sum-product bounds via Holder's inequality. (Weighted mini-buckets)
Week 8	2/25	William: Daniel Lowd, University of Oregon; Pedro Domingos. Approximate Inference by Compilation to Arithmetic Circuits. University of Washington.Slides Teague: Tomas Werner. A Linear Programming Approach to Max-sum Problem: A Review. slides, paper, extended paper.
Week 9	3/4	Jenny Lam: Jim C. Huang, Nebojsa Jojic and Christopher Meek. Exact inference and learning for cumulative distribution functions on loopy graphs. Slides
Week 10	3/11