Title: Objective Bayesian Analysis

Speaker: Prof.James Berger

Duke University and the Statistical and Applied Mathematical Sciences

Institute

Time: Jan.7(Monday),10:00-11:00am,2008

Place: Rm115

Abstract: Objective Bayesian analysis has existed for over 240 years, and has been central to the development of statistics over that period, yet the common perception today is that Bayesian analysis is a subjectivist methodology. Some reasons for this and highlights of the objective Bayesian history will first be reviewed.

The benefits in utilizing the objective Bayesian approach will next be discussed, utilizing examples that range from medical diagnosis to determination of distances to Cepheid variable stars. If time permits, some discussion will be given of the Bayesian/frequentist unification that is possible through the objective Bayesian approach, and the role of frequentist concepts such as coverage and admissibility in the choice of objective priors.