Probabilistic Graphical Models  10-708, Fall 2006 School of Computer Science, Carnegie-MellonUniversity

---

Syllabus and Course Schedule

Module	Material covered	Online material and links	Dates and Instructor
Module 0: introduction (1 Lectures)		Sep 13: Lecture Slides Sep 13: Annotated Lecture Slides JavaBayes Applet	Sep 13: Carlos Guestrin
Module 1: Fundamentals of Bayesian Networks (directed graphical models): representation, semantics, learning and inference (9 Lectures)	Representation Semantics MLE parameter learning Structure learning for BNs - complete data Constraint-based Chow-Liu Fixed-order Structure search Exact inference Variable elimination  Junction trees Context-specific independence	Sep 15: Lecture Slides Sep 15: Annotated Lecture Slides Sep 20: Lecture Slides Sep 20: Annotated Lecture Slides Sep 22: Lecture Slides Sep 22: Annotated Lecture Slides Sep 27: Lecture Slides Sep 27: Annotated Lecture Slides Sep 29: Lecture Slides Sep 29: Annotated Lecture Slides Oct 4: Lecture Slides Oct 4: Annotated Lecture Slides Oct 6: Lecture Slides Oct 6: Annotated Lecture Slides Oct 11: Lecture Slides Oct 11: Annotated Lecture Slides Oct 13: Lecture Slides Oct 13: Annotated Lecture Slides Oct 18: Lecture Slides Oct 18: Annotated Lecture Slides	Sep 15: Carlos Guestrin Sep 20: Carlos Guestrin Sep 22: Carlos Guestrin Sep 27: Carlos Guestrin Sep 29: Carlos Guestrin Oct 4: Carlos Guestrin Oct 6: Carlos Guestrin Oct 11: Carlos Guestrin Oct 13: Carlos Guestrin Oct 18: Carlos Guestrin Oct 20: NO CLASS
Module 2: Undirected Graphical Models, Markov random fields, Factor Graphs (1 Lectures)	Undirected models  - Representation Factor graphs - unifying representation Exponential family	Oct 25: Lecture Slides Oct 25: Annotated Lecture Slides	Oct 25: Carlos Guestrin
Module 3: Approximate inference  (3 Lectures)	Sampling importance sampling MCMC, Gibbs Variational inference Loopy belief propagation Generalized belief propagation Kikuchi	Oct 27: Lecture Slides Oct 27: Annotated Lecture Slides Yedidia et al. (2005) have a nice overview of free energy approximations and GBP.  For perhaps a slightly more gentle tutorial, see Yedidia et al. (2001). Nov 1: Lecture Slides Nov 1: Annotated Lecture Slides Nov 3: Lecture Slides Nov 3: Annotated Lecture Slides Nov 8: Lecture Slides Nov 8: Annotated Lecture Slides	Oct 27: Carlos Guestrin Nov 1: Carlos Guestrin Nov 3: Ajit Singh
Module 4: Learning revisited (2 Lectures)	Parameter estimation in BNs with missing data EM Gradient descent Structure learning for BNs - missing data Learning undirected graphical models Gradient algorithms IPF for tabular MRFs Structure Learning	Nov 10: Lecture Slides Nov 10: Annotated Lecture Slides Nov 15: Lecture Slides Nov 15: Annotated Lecture Slides	Nov 8:Carlos Guestrin Nov 10:Carlos Guestrin Nov 15:Carlos Guestrin
Module 5: Gaussian and hybrid models (1 Lecture)	Multivariate Gaussians Gaussian graphical models Inference in Gaussian models Hybrid models discrete and continuous variables	Nov 17: Lecture Slides Nov 17: Annotated Lecture Slides	Nov 17:Carlos Guestrin
Module 6: Temporal models (1 Lecture)	Hidden Markov Models Representation Inference Forwards-Backwards Viterbi Baum-Welch Kalman filter representation linearization switching Kalman filter assumed density filtering DBNs Representation Inference BK Relational probabilistic models	Nov 20: Lecture Slides Nov 20: Annotated Lecture Slides Nov 27: Lecture Slides Nov 27: Annotated Lecture Slides Nov 29: Causality Slides Nov 29: Teeth Example Slides Dec 1: Lecture Slides Dec 1: Annotated Lecture Slides	Special time: Nov 20 - 5:30-7pm, Wean 4615A: Carlos Guestrin Special time: Nov 27 - 5:30-7pm, Wean 4615A: Carlos Guestrin
Module 7: Advanced Topics  (2 Lectures)	Causality Relational probabilistic models Template models		Nov 22: NO CLASS  Nov 24: NO CLASS Nov 29: Guest lecturer: Richard Scheines on causality!  Dec 1: Carlos Guestrin
Project poster session			Dec 1, 3-6pm
Final Exam	All material thus far		Out Dec 1. Due Dec 15 by 2pm

Recitation Schedule

Date	Time	Place	Topic	Instructor
Thursday 9/14/06	5-6:30pm	Wean 4615A	Review of basic probabilities	Andreas Krause
Monday 9/18/06	5:30-7pm	Wean 4615A	Matlab Tutorial	Jure Leskovec
Thursday 9/21/06	5-6:30pm	Wean 4615A	BN Semantics	Anna Goldenberg
Thursday 9/28/06	5-6:30pm	Wean 4615A	P-maps, Bayesian parameter learning	Khalid
Thursday 10/5/06	5-6:30pm	Wean 4615A	Structure learning	Ajit
Thursday 10/12/06	5-6:30pm	Wean 4615A	Variable Elimination	Khalid
Monday 10/16/06	5:30-7pm	Wean 4615A	Context-Specific Independence Slides Annotated Slides	Carlos
Thursday 10/19/06	5-6:30pm	Wean 4615A		Ajit
Thursday 10/26/06	5-6:30pm	Wean 4615A	Pairwise Markov Networks	Khalid
Thursday 11/2/06	5-6:30pm	Wean 4615A	Mean Field Variational Inference	Khalid
Monday 11/6/06	5:30-7pm	Wean 4615A	Dirichlet Process Mixtures	Khalid
Thursday 11/9/06	5:00-6:30pm	Wean 4615A	Loopy BP	Ajit
Monday 11/13/06	5:30-6:30pm	Wean 4615A	Relaxation Methods for Inference	Pradeep Ravikumar