STAT471-12S2 (C) Semester Two 2012
Special Topic in Statistics

Description

Special Topic in 2012 - Nonparametric Bayesian Inference

The goal of nonparametric inference is flexibility in modelling and robustness of the resulting inference. In Bayesian nonparametrics, this is achieved through the use of random processes as priors on infinite-dimensional function spaces.

This course looks at some of these priors together with the computational methods for using them. Their use will be illustrated through appications such as density estimation, regression, classification and clustering.

Students should have a sound mathematical background and a good foundation in Statistics and Probability, at least up to the level of STAT213 or STAT214. They should preferably also have taken, or are concurrently taking, STAT314 or STAT461 or a similar course in parametric Bayesian statistics.

For a full list of 2012 Honours courses, please refer to the Department of Mathematics and Statistics Honours Booklet Mathematics and Statistics Honours Booklet

For further information see Mathematics and Statistics.