Some Bayesian predictive approaches to model selection

Mukhopadhyay, Nitai ; Ghosh, Jayanta K. ; Berger, James O. (2005) Some Bayesian predictive approaches to model selection Statistics & Probability Letters, 73 (4). pp. 369-379. ISSN 0167-7152

Full text not available from this repository.

Official URL: http://linkinghub.elsevier.com/retrieve/pii/S01677...

Related URL: http://dx.doi.org/10.1016/j.spl.2005.04.012

Abstract

A variety of pseudo-Bayes factors have been proposed, based on using part of the data to update an improper prior, and using the remainder of the data to compute the Bayes factor. A number of these approaches are of a bootstrap or cross-validation nature, with some type of average being taken over the data used for updating. Asymptotic characteristics of a number of these pseudo-Bayes factors are discussed, and it is shown how many behave quite differently from ordinary Bayes factors. It is also shown that arguments of predictive optimality, based on simply inserting the empirical distribution in place of the 'true predictive distribution', can be misleading; the particular example of this that is studied is the argument given in Bernardo and Smith [1994. Bayesian Theory. Wiley, Chichester] to the effect that the geometric intrinsic Bayes factor has an optimal predictive property.

Item Type:Article
Source:Copyright of this article belongs to Elsevier Science.
Keywords:Bayes Factor; Training Sample
ID Code:22533
Deposited On:24 Nov 2010 08:23
Last Modified:02 Jun 2011 06:42

Repository Staff Only: item control page