Marginal Likelihood From the Metropolis–Hastings Output
- 1 March 2001
- journal article
- Published by Taylor & Francis Ltd in Journal of the American Statistical Association
- Vol. 96 (453), 270-281
- https://doi.org/10.1198/016214501750332848
Abstract
This article provides a framework for estimating the marginal likelihood for the purpose of Bayesian model comparisons. The approach extends and completes the method presented in Chib (1995) by overcoming the problems associated with the presence of intractable full conditional densities. The proposed method is developed in the context of MCMC chains produced by the Metropolis–Hastings algorithm, whose building blocks are used both for sampling and marginal likelihood estimation, thus economizing on prerun tuning effort and programming. Experiments involving the logit model for binary data, hierarchical random effects model for clustered Gaussian data, Poisson regression model for clustered count data, and the multivariate probit model for correlated binary data, are used to illustrate the performance and implementation of the method. These examples demonstrate that the method is practical and widely applicable.Keywords
This publication has 17 references indexed in Scilit:
- On MCMC sampling in hierarchical longitudinal modelsStatistics and Computing, 1999
- Posterior simulation and Bayes factors in panel count data modelsJournal of Econometrics, 1998
- Analysis of multivariate probit modelsBiometrika, 1998
- Computing Bayes Factors by Combining Simulation and Asymptotic ApproximationsJournal of the American Statistical Association, 1997
- On Monte Carlo methods for estimating ratios of normalizing constantsThe Annals of Statistics, 1997
- Reversible jump Markov chain Monte Carlo computation and Bayesian model determinationBiometrika, 1995
- Marginal Likelihood from the Gibbs OutputJournal of the American Statistical Association, 1995
- Understanding the Metropolis-Hastings AlgorithmThe American Statistician, 1995
- Bayesian Analysis of Binary and Polychotomous Response DataJournal of the American Statistical Association, 1993
- Sampling-Based Approaches to Calculating Marginal DensitiesJournal of the American Statistical Association, 1990