Bayesian Statistics: 3 hours

Catalog description

(Prerequisite: ST 4113 ). Three hours lecture. This course introduces the Bayesian approach to statistical inference, emphasizing both theory and computation. Beginning with simple models based on normal and binomial distributions, the course illustrates concepts of conjugate and non-informative priors for single- and multi-parameter models. Topics include Bayes’ theorem, prior and posterior distributions, hypothesis testing, Bayesian regression, and model comparison using Bayes factors. Computational methods such as Monte Carlo simulation and Gibbs sampling algorithms are presented with an emphasis on implementation in R and BUGS