Bayesian and causal Bayesian Networks

Undergraduate

Representation (factorizing joint probabilistic distributions, exploiting probabilistic independence, d- separation, I-map, naive Bayes); parameter learning (maximum likelihood estimation, Bayesian parameter estimation); structure learning (constrained and score-based approaches, Bayesian model averaging); learning with incomplete data (parameter and structure learning, learning with hidden variables); causal Bayesian networks (intervention, SCMs, inference, learning); Inference. Use of R or Python. Course Format: Lectures, 3 hours per week; lab/tutorial, 1 hour per week. Prerequisite(s): STAT 3P85 or the permission of the instructor.