Many problems in fields such as artificial intelligence, statistics, computer vision, natural language processing, and computational biology can be viewed as the search for a coherent global conclusion from local information. The probabilistic graphical models (PGMs) framework provides a unified approach for solving this wide range of problems, enabling efficient inference, decision-making, and learning in systems with a large number of attributes and huge datasets.

This course will provide a strong foundation for applying PGMs to complex real-world problems, as well as addressing core research topics in graphical models. Students will learn to construct both Bayesian and Markov networks, perform inference, and apply learning algorithms. Toward the end of the course, we will also explore modern probabilistic AI techniques, including generative models and large language models (LLMs), emphasizing how PGMs underpin many state-of-the-art AI systems.

  • Time: Tuesday/Thursday 11:00am-12:15pm
  • Location: Microbial Sciences Building 1520
  • Discussion: Canvas
  • HW submission: Canvas
  • Contact: Students should ask all course-related questions on Canvas, where you will also find announcements. Individual enquires can be directed to TA/instructor emails.