Marginal likelihood for a class of Bayesian generalized linear models

Balgobin Nandram, Hyunjoong Kim

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

The Bayes factor has become an important tool for model selection. The marginal likelihoods are also important because they can be used to rank the models. In fact, the Bayes factor is the ratio of the marginal likelihoods for two models with proper prior densities. We discuss the marginal likelihood for a class of generalized linear models used in small area estimation for mortality data analysis. Computation in these models is intensive and requires the implementation of Markov chain Monte Carlo (MCMC) methods. A sophisticated method for computing the marginal likelihoods for generalized linear models using reduced Metropolis-Hastings (M-H) samplers has recently been introduced. Also, a much simpler method that uses the Laplace approximation has been proposed. Our method lies between these two in simplicity, and it uses importance sampling via a simple output analysis from a MCMC sampler. We also show that the new method can be approximated without using the MCMC sampler, We illustrate our methods for two Poisson regression models which have been used for mapping mortality data, and we compare them with the methods based on reduced M-H samplers and the Laplace approximation. The method based on the Laplace approximation gives smaller marginal log-likelihoods than the other three methods, but they select the same model, We use a simulation study in which the marginal likelihood is used to investigate the improvement in the goodness of fit of one hierarchical Poisson regression model over the other.

Original languageEnglish
Pages (from-to)319-340
Number of pages22
JournalJournal of Statistical Computation and Simulation
Volume72
Issue number4
DOIs
Publication statusPublished - 2002 Dec 1

Fingerprint

Marginal Likelihood
Generalized Linear Model
Laplace Approximation
Metropolis-Hastings
Poisson Regression
Markov processes
Bayes Factor
Poisson Model
Markov Chain Monte Carlo
Mortality
Regression Model
Small Area Estimation
Class
Marginal likelihood
Generalized linear model
Importance sampling
Importance Sampling
Markov Chain Monte Carlo Methods
Goodness of fit
Model Selection

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Modelling and Simulation
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

@article{c9f5e48848d04d66840ae5a051457c4b,
title = "Marginal likelihood for a class of Bayesian generalized linear models",
abstract = "The Bayes factor has become an important tool for model selection. The marginal likelihoods are also important because they can be used to rank the models. In fact, the Bayes factor is the ratio of the marginal likelihoods for two models with proper prior densities. We discuss the marginal likelihood for a class of generalized linear models used in small area estimation for mortality data analysis. Computation in these models is intensive and requires the implementation of Markov chain Monte Carlo (MCMC) methods. A sophisticated method for computing the marginal likelihoods for generalized linear models using reduced Metropolis-Hastings (M-H) samplers has recently been introduced. Also, a much simpler method that uses the Laplace approximation has been proposed. Our method lies between these two in simplicity, and it uses importance sampling via a simple output analysis from a MCMC sampler. We also show that the new method can be approximated without using the MCMC sampler, We illustrate our methods for two Poisson regression models which have been used for mapping mortality data, and we compare them with the methods based on reduced M-H samplers and the Laplace approximation. The method based on the Laplace approximation gives smaller marginal log-likelihoods than the other three methods, but they select the same model, We use a simulation study in which the marginal likelihood is used to investigate the improvement in the goodness of fit of one hierarchical Poisson regression model over the other.",
author = "Balgobin Nandram and Hyunjoong Kim",
year = "2002",
month = "12",
day = "1",
doi = "10.1080/00949650212842",
language = "English",
volume = "72",
pages = "319--340",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

Marginal likelihood for a class of Bayesian generalized linear models. / Nandram, Balgobin; Kim, Hyunjoong.

In: Journal of Statistical Computation and Simulation, Vol. 72, No. 4, 01.12.2002, p. 319-340.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Marginal likelihood for a class of Bayesian generalized linear models

AU - Nandram, Balgobin

AU - Kim, Hyunjoong

PY - 2002/12/1

Y1 - 2002/12/1

N2 - The Bayes factor has become an important tool for model selection. The marginal likelihoods are also important because they can be used to rank the models. In fact, the Bayes factor is the ratio of the marginal likelihoods for two models with proper prior densities. We discuss the marginal likelihood for a class of generalized linear models used in small area estimation for mortality data analysis. Computation in these models is intensive and requires the implementation of Markov chain Monte Carlo (MCMC) methods. A sophisticated method for computing the marginal likelihoods for generalized linear models using reduced Metropolis-Hastings (M-H) samplers has recently been introduced. Also, a much simpler method that uses the Laplace approximation has been proposed. Our method lies between these two in simplicity, and it uses importance sampling via a simple output analysis from a MCMC sampler. We also show that the new method can be approximated without using the MCMC sampler, We illustrate our methods for two Poisson regression models which have been used for mapping mortality data, and we compare them with the methods based on reduced M-H samplers and the Laplace approximation. The method based on the Laplace approximation gives smaller marginal log-likelihoods than the other three methods, but they select the same model, We use a simulation study in which the marginal likelihood is used to investigate the improvement in the goodness of fit of one hierarchical Poisson regression model over the other.

AB - The Bayes factor has become an important tool for model selection. The marginal likelihoods are also important because they can be used to rank the models. In fact, the Bayes factor is the ratio of the marginal likelihoods for two models with proper prior densities. We discuss the marginal likelihood for a class of generalized linear models used in small area estimation for mortality data analysis. Computation in these models is intensive and requires the implementation of Markov chain Monte Carlo (MCMC) methods. A sophisticated method for computing the marginal likelihoods for generalized linear models using reduced Metropolis-Hastings (M-H) samplers has recently been introduced. Also, a much simpler method that uses the Laplace approximation has been proposed. Our method lies between these two in simplicity, and it uses importance sampling via a simple output analysis from a MCMC sampler. We also show that the new method can be approximated without using the MCMC sampler, We illustrate our methods for two Poisson regression models which have been used for mapping mortality data, and we compare them with the methods based on reduced M-H samplers and the Laplace approximation. The method based on the Laplace approximation gives smaller marginal log-likelihoods than the other three methods, but they select the same model, We use a simulation study in which the marginal likelihood is used to investigate the improvement in the goodness of fit of one hierarchical Poisson regression model over the other.

UR - http://www.scopus.com/inward/record.url?scp=0038826034&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0038826034&partnerID=8YFLogxK

U2 - 10.1080/00949650212842

DO - 10.1080/00949650212842

M3 - Article

VL - 72

SP - 319

EP - 340

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 4

ER -