41505 - Bayesian Statistical Inference

Learning outcomes

Introductory course in Bayesian Statistics. Parametric estimation, predictive inference and hypothesis testing are revisited according to the Bayesian paradigm. Numerical methods for solving estimation problems are introduced.

Course contents

Classical statistical inference. Inference based on the likelihood principle. Bayesian inference. Subjective probability, conditional probability.
The posterior distribution for Bernoulli trials and binomial responses considering different prior distributions.
The role of natural conjugate distributions in Bayesian inference. Improper prior distributions in Bayesian inference.
Interval estimation in Bayesian inference.
Introduction to hypothesis testing in Bayesian inference. The Bayes factor.
Hierarchical Bayesian models.
Conditional independence and graphical representation of hierarchical models.
Introduction to McMC methods. The WinBUGS software for Bayesian model estimation.
Examples of Bayesian model estimation with the WinBUGS software.

Readings/Bibliography

Peter D. Hoff (2009) A First Course in Bayesian Statistical Methods, Springer

D. Michele Cifarelli, Pietro Muliere (1989) Statistica Bayesiana: appunti ad uso degli studenti, Iuculano.
Peter M. Lee (1997) Bayesian statistics: an introduction, Arnold
Peter Congdon (2001) Bayesian Statistical modelling, Wiley

Teaching methods

lectures and practical classes with WinBugs

Assessment methods

Test in the computer room using WInBugs and oral test

Teaching tools

WinBugs sessions

Office hours

See the website of Daniela Cocchi