Home > Courses
Teaching
Dr. Margaret Chen, Stanford Computer Science
| CS 228: Probabilistic Graphical Models |
Autumn Quarter 1997 — MW 1:15-2:30, Gates B01
Prerequisites: CS 109 (probability), CS 161 (algorithms), linear algebra
Office Hours: Wednesday 3:00-5:00 PM, Gates 156
TA: David Park (park@cs.stanford.edu), OH: Thursday 2:00-4:00, Gates 164
|
| Course Description |
This course covers the fundamentals of probabilistic graphical models, including Bayesian networks, Markov random fields, and their applications to reasoning under uncertainty. Topics include exact and approximate inference, parameter estimation, structure learning, and connections to decision theory.
New this year: Two lectures on variational methods and the wake-sleep algorithm for learning in deep belief networks. Also covering recent work on influence diagrams for sequential decision-making.
|
| Tentative Schedule |
| Week 1 | Probability review, conditional independence |
| Week 2 | Bayesian networks: representation |
| Week 3 | Exact inference: variable elimination, junction tree |
| Week 4 | Approximate inference: sampling methods |
| Week 5 | Variational methods, mean field |
| Week 6 | Midterm Exam |
| Week 7 | Parameter estimation, EM algorithm |
| Week 8 | Structure learning, scoring criteria |
| Week 9 | Markov random fields, undirected models |
| Week 10 | Decision networks, influence diagrams, project presentations |
|
| Textbook |
|
Probabilistic Reasoning in Intelligent Systems by Judea Pearl (Morgan Kaufmann, 1988). Supplementary readings will be distributed as photocopies.
|
| CS 371: Seminar on Knowledge Representation |
Winter Quarter 1998 — F 2:15-3:30, Gates 219
Format: Student-led paper discussions, 2 units S/NC
A reading seminar exploring recent advances in knowledge representation and automated reasoning. Topics include description logics, nonmonotonic reasoning, belief revision, and the frame problem. We will read papers from KR-97, AAAI-97, and IJCAI-97.
Interested? Email me for the reading list. Enrollment limited to 15.
|
Previously Taught:
• CS 221: Artificial Intelligence (Spring '96, '97)
• CS 228: Probabilistic Graphical Models (Autumn '95, '96)
• CS 399: Independent Study — topics vary
|
Home | Research | Courses | Publications | Hot Links
|