CS673 - Probabilistic Reasoning in AI

Syllabus

• Introduction: Basic probability notations, Axioms of probability, Probability distributions, Bayes theorem.
  

• Bayesian Networks: Semantics of Bayesian networks, Exact inferencing - enumeration, variable elimination, junction-tree Approximate inferencing - sampling methods, Markov chain Monte Carlo methods, variational methods, particle filtering.
  

• Reasoning over time: inference in temporal models, hidden markov models, Kalman filters, dynamic Bayesian networks, efficient representations of CPTs.
  

• Decision making under uncertainity: Beliefs and utility, Utility theory, utility functions, risk modeling, decision networks, Value of Information.
  

• Decision theoretic planning: Sequential decision making, Markov decision processes. Value iteration, policy iteration, Real-time dynamic programming, multi-agent planning: game theory, reinforcement learning, rolls outs.
  

Department of Computer Science and Engineering, Indian Institute of Technology Madras, Chennai 600036
Phone: +91-44-22574350 Fax: +91-44- 22574352 email: csoffice@iitm.ac.in