Laboratory of Statistical Artificial Intelligence and Machine Learning

Probabilistic Graphical Models Fall 2025

Description

Probabilistic graphical models are a useful tool for representing domain knowledge using probability distributions. They have a wide range of applications in fields such as machine learning, computer vision, and natural language processing, The combination of graph theory and probability theory provides a rich framework for modeling domain knowledge consisting of a large number of random variables with interdependencies. This course will provide a thorough overview of the topic, including the key formalisms and techniques used to build, predict, and make decisions under uncertainty.

Lecture Timings (Tentative)

Tuesday 3.30-4.45pm
Thursday 2.00-3.15pm

Reference Material

Primary Textbook : Probabilistic Graphical Models: Principles and Techniques, Daphne Koller and Nir Friedman, MIT Press
Other Reference Books: Modeling and Reasoning with Bayesian Networks, Adnan Darwiche, Cambridge University Press.

Academic integrity

Students enrolled in this course are expected to exhibit a strong desire to learn, rather than just fulfilling a requirement for their degree. Engaging in discussions that help students better understand concepts or problems is encouraged. However, all submitted work must be original. Plagiarism, including copying from the internet, textbooks, or any other source for which the student does not hold the copyright, as well as sharing code with other students, will not be tolerated and will result in strict disciplinary action, including a failing grade in the course. If you have any questions about this policy, please contact the instructor. All academic integrity violations will be handled in accordance with institute regulations.

Grading Policy

Quiz Two Quizzes (each worth 20) that will account for 40 of the overall grade.

Assignments Two-Three Assignments that will account for 10% of the grade.

Paper Presentations Research paper presentation that will account for 10% of the overall grade

Project Semester long project for 40% of the overall grade.

Attendance

This course follows the attendance criteria mandated by the institute.

Course Topics

Introduction
Bayesian Networks
Undirected Graphical Models
Probabilistic Circuits
Exact Inference
Approximate Infeerence
Parameter Learning (MLE and EM)
Structure Learning
HMM
Variational Inference
Deep Generative Models
Deep Generative Models
Advanced Topics

Lecture Material

Student enrolled in the course can access the lecture material from Google Classroom Link