Open Access
Translator Disclaimer
2012 A survey of Bayesian predictive methods for model assessment, selection and comparison
Aki Vehtari, Janne Ojanen
Statist. Surv. 6: 142-228 (2012). DOI: 10.1214/12-SS102


To date, several methods exist in the statistical literature for model assessment, which purport themselves specifically as Bayesian predictive methods. The decision theoretic assumptions on which these methods are based are not always clearly stated in the original articles, however. The aim of this survey is to provide a unified review of Bayesian predictive model assessment and selection methods, and of methods closely related to them. We review the various assumptions that are made in this context and discuss the connections between different approaches, with an emphasis on how each method approximates the expected utility of using a Bayesian model for the purpose of predicting future data.


Download Citation

Aki Vehtari. Janne Ojanen. "A survey of Bayesian predictive methods for model assessment, selection and comparison." Statist. Surv. 6 142 - 228, 2012.


Published: 2012
First available in Project Euclid: 27 December 2012

zbMATH: 1302.62011
MathSciNet: MR3011074
Digital Object Identifier: 10.1214/12-SS102

Primary: 62-02
Secondary: 62C10

Keywords: Bayesian , cross-validation , decision theory , Expected utility , information criteria , model assessment , Model selection , predictive

Rights: Copyright © 2012 The author, under a Creative Commons Attribution License


Vol.6 • 2012
Back to Top