Stat 460 - Bayesian Statistics
Lectures: Tuesdays 11:30-1:20 in AQ5018 and
Thursdays 11:30 in WMC3220,
Tutorial Fridays 10:30 in SECB1010
Office Hours: Thursdays after class in K10539
or by appointment (send emails only in case of emergencies)
Reference Books (my course notes will be provided):
- Carlin and Louis - Bayes and Empirical Bayes Methods for Data
Analysis
- Gelman, Carlin, Stern and Rubin - Bayesian Data Analysis
- Bernardo and Smith - Bayesian Theory
- Gilks, Richardson and Spiegelhalter - Markov Chain Monte Carlo
in Practice
- Lee - Bayesian Statistics: An Introduction, Fourth Edition,
Marking Scheme:
- Participation 15 marks
- Assignments 20 marks
- Midterm - 20 marks - Tuesday Feb 12
- Final Exam - 45 marks - Tuesday April 16 12:00 - 3:00pm
Course Outline:
1. The Basics
- the Bayesian paradigm
- views of probability
2. Comparative Inference
- arguments for/against the Bayesian approach
- examples of crazy frequentist procedures
3. Priors
- conjugate priors
- prior elicitation
- reference priors
- improper priors
- discrete mass priors
4. Computation
- quadrature
- importance sampling
- Gibbs sampling
- other MCMC procedures
5. Other Topics
- testing via Bayes factors
- interval and point estimation
- elementary decision theory
- hierarchical models
- Dirichlet process
- WinBUGS
6. Applications