Prior-based Bayesian information criterion

Published

Journal Article

© 2019, © East China Normal University 2019. We present a new approach to model selection and Bayes factor determination, based on Laplace expansions (as in BIC), which we call Prior-based Bayes Information Criterion (PBIC). In this approach, the Laplace expansion is only done with the likelihood function, and then a suitable prior distribution is chosen to allow exact computation of the (approximate) marginal likelihood arising from the Laplace approximation and the prior. The result is a closed-form expression similar to BIC, but now involves a term arising from the prior distribution (which BIC ignores) and also incorporates the idea that different parameters can have different effective sample sizes (whereas BIC only allows one overall sample size n). We also consider a modification of PBIC which is more favourable to complex models.

Full Text

Duke Authors

Cited Authors

  • Bayarri, MJ; Berger, JO; Jang, W; Ray, S; Pericchi, LR; Visser, I

Published Date

  • January 2, 2019

Published In

  • Statistical Theory and Related Fields

Volume / Issue

  • 3 / 1

Start / End Page

  • 2 - 13

Electronic International Standard Serial Number (EISSN)

  • 2475-4277

International Standard Serial Number (ISSN)

  • 2475-4269

Digital Object Identifier (DOI)

  • 10.1080/24754269.2019.1582126

Citation Source

  • Scopus