Probabilistic graphical models provide a convenient framework for modeling of joint distributions by utilizing graphs to represent the dependence among the variables. The course introduces several such frameworks, including Bayesian (belief) networks, Markov random fields, and covers topics related to representation, exact and approximate inference, and parameter and structure estimation in models for high-dimensional data.
47822 STAT 59800 - SK1
Monday, Wednesday, and Friday, 9:30am-10:20am, 103 University Hall
Monday 1:30-2:30pm in HAAS 118 or by appointment
Probabilistic Graphical Models: Principles and Techniques by Daphne Koller and Nir Friedman, The MIT Press, 2009 (official errata, our additional errata) required
A basic course on probability (e.g., STAT 416/516); programming experience (e.g., STAT 598G); course on linear algebra recommended; introduction course in machine learning (STAT 598A/CS 578) helpful but not required. If not sure, please discuss with the instructor as early as possible.
For grading, policies, and what to expect, please read the syllabus.