Bayesian Measures of Model Complexity and Fit
Top Cited Papers
Open Access
- 1 October 2002
- journal article
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 64 (4), 583-639
- https://doi.org/10.1111/1467-9868.00353
Abstract
Summary. We consider the problem of comparing complex hierarchical models in which the number of parameters is not clearly defined. Using an information theoretic argument we derive a measure pD for the effective number of parameters in a model as the difference between the posterior mean of the deviance and the deviance at the posterior means of the parameters of interest. In general pD approximately corresponds to the trace of the product of Fisher's information and the posterior covariance, which in normal models is the trace of the ‘hat’ matrix projecting observations onto fitted values. Its properties in exponential families are explored. The posterior mean deviance is suggested as a Bayesian measure of fit or adequacy, and the contributions of individual observations to the fit and complexity can give rise to a diagnostic plot of deviance residuals against leverages. Adding pD to the posterior mean deviance gives a deviance information criterion for comparing models, which is related to other information criteria and has an approximate decision theoretic justification. The procedure is illustrated in some examples, and comparisons are drawn with alternative Bayesian and classical proposals. Throughout it is emphasized that the quantities required are trivial to compute in a Markov chain Monte Carlo analysis.Keywords
Funding Information
- Medical Research Council (G9803841)
- National Institute of Allergy and Infectious Diseases (1-R01-AI41966)
This publication has 78 references indexed in Scilit:
- Computational and Inferential Difficulties with Mixture Posterior DistributionsJournal of the American Statistical Association, 2000
- Reference Bayesian Methods for Generalized Linear Mixed ModelsJournal of the American Statistical Association, 2000
- Nonconjugate Bayesian Estimation of Covariance Matrices and its Use in Hierarchical ModelsJournal of the American Statistical Association, 1999
- On Measuring and Correcting the Effects of Data Mining and Model SelectionJournal of the American Statistical Association, 1998
- Bayes FactorsJournal of the American Statistical Association, 1995
- Parameterizations for Natural Exponential Families with Quadratic Variance FunctionsJournal of the American Statistical Association, 1994
- Generalized Linear Models with Random Effects; a Gibbs Sampling ApproachJournal of the American Statistical Association, 1991
- How Biased is the Apparent Error Rate of a Prediction Rule?Journal of the American Statistical Association, 1986
- Estimating the Dimension of a ModelThe Annals of Statistics, 1978
- Science and StatisticsJournal of the American Statistical Association, 1976