University of Newcastle upon Tyne

School of Mathematics and Statistics

Statistics Seminars 2005-2006

25 November 2005, L401, 2:15pm

Dr Jim Griffin, University of Warwick

Abstract

Bayesian nonparametric methods have generally been applied to data which are assumed to be independent and identically distributed (i.i.d.) from an unknown distribution. Applications have appeared in many areas including meta-analysis, econometric modelling and density estimation. This talk will discuss nonparametric priors when the i.i.d. assumption is not valid, either through modelling of the transition kernel of a Markov chain or through evolving (nonparametric) distributions. The ideas will be illustrated by applications to data.
 

Return to timetable for 2005-06 seminars

School of Mathematics and Statistics Homepage