Probabilistic Graphical Models

10-708, Fall 2021
DH 2302, Tue & Thurs 1:25PM - 2:45PM

Instructor Pradeep Ravikumar (pradeepr at cs dot cmu dot edu)
Teaching Assistants Swaminathan Gurumurthy (sgurumur at andrew dot cmu dot edu)
Kai-Ling Luo (kailingl at andrew dot cmu dot edu)
Xiang Si (xsi at andrew dot cmu dot edu)
Helen Zhou (hlzhou at andrew dot cmu dot edu)
Office Hours
Pradeep Ravikumar: Thursdays,3:40pm - 4:00pm,Zoom
For office hours of the TAs and Zoom links, please check Piazza.
Course details Syllabus. Project Instructions. Piazza. Gradescope.

Grading 75% Homeworks, 25% Project. For further details, please see the Syllabus.

Textbooks Lecture notes will be posted for each class, which will be largely self-contained. For further reference, we recommend the following textbooks:
  • KF: Probabilistic Graphical Models: Principles and Techniques, by Daphne Koller and Nir Friedman.
  • WJ: Graphical Models, Exponential Families, and Variational Inference, by Martin Wainwright, Michael Jordan


    		•
    Tentative Schedule
    Date Topic Readings Notes
    Aug 31 Introduction, the what, the why - -
    Sep 02 Undirected GMs, Markov Properties - -
    Sep 03 No recitation - -
    Sep 07 Directed GMs, Markov Properties - -
    Sep 09 Directed & Undirected GMs - -
    Sep 10 Recitation: Helen - -
    Sep 14 Chain GMs, Markov Properties - HW1 out
    Sep 16 DGMs and UGMs: Parameterization - -
    Sep 17 No recitation - -
    Sep 21 Recap, Applications - -
    Sep 23 Causal GMs: Intro, Structural Causal Models - -
    Sep 24 Recitation: Swami - -
    Sep 28 Causal GMs: Treatment Effects - HW1 due, HW2 out
    Sep 30 Exact Inference: Variable Elimination - -
    Oct 01 No recitation - -
    Oct 05 Exact Inference: Junction Trees, Message Passing - -
    Oct 07 MAP Estimation - -
    Oct 08 Recitation: Helen - -
    Oct 12 Inference: exponential family & optimization viewpoint - HW2 due, HW3 Out
    Oct 14 No Class: Midsemester Break - -
    Oct 15 No recitation - -
    Oct 19 Variational Inference: Sum-Product - Project proposal due
    Oct 21 Variational Inference: Upper & Lower Bounds - -
    Oct 22 Recitation: Swami - -
    Oct 26 Latent Variable Models: Exponential Families, Variational EM - HW3 due, HW4 out
    Oct 28 Latent Variable Models: Building Blocks, Examples - -
    Oct 29 No recitation - -
    Nov 02 Dynamic LVMs: HMMs, CRFs, State Space Models - -
    Nov 04 Hierarchical LVMs, Nonparametric Bayesian Models - -
    Nov 05 Recitation: Helen - -
    Nov 09 Neural PGMs, Deep Generative Models - HW4 due; HW5 out
    Nov 11 Neural PGMs, Deep Generative Models - -
    Nov 12 No recitation - -
    Nov 16 Learning PGMs - Project Midway Report due
    Nov 18 Approximate Inference: Sampling - -
    Nov 19 Recitation: Swami - -
    Nov 23 Approximate Inference: Sampling HW5 due
    Nov 25 No Class: Thanksgiving - -
    Nov 30 Class Project Presentations - -
    Dec 02 Class Project Presentations - -
    Dec 07 No class - Final report due