ADAPTIVE BAYESIAN CRITERIA IN VARIABLE SELECTION FOR GENERALIZED LINEAR MODELS

For the problem of variable selection in generalized linear models, we develop various adaptive Bayesian criteria. Using a hierarchical mixture setup for model uncertainty, combined with an integrated Laplace approximation, we derive Empirical Bayes and Fully Bayes criteria that can be computed easily and quickly. The performance of these criteria is assessed via simulation and compared to other criteria such as AIC and BIC on normal, logistic and Poisson regression model classes. A Fully Bayes criterion based on a restricted region hyperprior seems to be the most promising. Finally, our criteria are illustrated and compared with competitors on a data example.

[1]  Wen Cui.  Variable selection: empirical Bayes vs. fully Bayes , 2002 .

[2]  Bent Jørgensen,et al.  Exponential Dispersion Models and Extensions: A Review , 1992 .

[3]  Xinlei Wang,et al.  Bayesian variable selection for GLM , 2002 .

[4]  J. Ibrahim,et al.  Prior elicitation, variable selection and Bayesian computation for logistic regression models , 1999 .

[5]  Thomas S. Shively,et al.  Variable Selection : Empirical Bayes vs . Fully Bayes , 2002 .

[6]  Adrian E. Raftery,et al.  Model Selection for Generalized Linear Models via GLIB, with Application to Epidemiology , 1993 .

[7]  R. Kohn,et al.  Nonparametric regression using Bayesian variable selection , 1996 .

[8]  N. Bleistein,et al.  Asymptotic Expansions of Integrals , 1975 .

[9]  A. Cameron,et al.  A Microeconometric Model of the Demand for Health Care and Health Insurance in Australia , 1988 .

[10]  M. Schemper,et al.  A solution to the problem of separation in logistic regression , 2002, Statistics in medicine.

[11]  Edward I. George,et al.  Two Approaches to Bayesian Model Selection with Applications , 1996 .

[12]  P. Dellaportas,et al.  Bayesian variable selection using the Gibbs sampler , 2000 .

[13]  A. Raftery.  Approximate Bayes factors and accounting for model uncertainty in generalised linear models , 1996 .

[14]  Dean Phillips Foster,et al.  Calibration and Empirical Bayes Variable Selection , 1997 .

[15]  李幼升,et al.  Ph , 1989 .

[16]  D. Freedman.  A Note on Screening Regression Equations , 1983 .

[17]  Tze Leung Lai,et al.  A hybrid estimator in nonlinear and generalised linear mixed effects models , 2003 .

[18]  Purushottam W. Laud,et al.  Predictive Variable Selection in Generalized Linear Models , 2002 .

[19]  P. McCullagh,et al.  Generalized Linear Models , 1984 .

[20]  L. Wasserman,et al.  A Reference Bayesian Test for Nested Hypotheses and its Relationship to the Schwarz Criterion , 1995 .

[21]  P. Dellaportas,et al.  Markov chain Monte Carlo model determination for hierarchical and graphical log-linear models , 1999 .

[22]  Joseph G. Ibrahim,et al.  BAYESIAN VARIABLE SELECTION FOR TIME SERIES COUNT DATA , 2000 .

[23]  Petros Dellaportas,et al.  On Bayesian model and variable selection using MCMC , 2002, Stat. Comput..

[24]  Pravin K. Trivedi,et al.  Regression Analysis of Count Data: Preface , 1998 .

[25]  L. Tierney,et al.  The validity of posterior expansions based on Laplace''s method , 1990 .

[26]  A. Cameron,et al.  Econometric models based on count data. Comparisons and applications of some estimators and tests , 1986 .

[27]  P. Dellaportas,et al.  Bayesian variable and link determination for generalised linear models , 2003 .

[28]  A. P. Dawid,et al.  Bayesian Model Averaging and Model Search Strategies , 2007 .