Consistency of hyper-$g$-prior-based Bayesian variable selection for generalized linear models

We study the consistency of a Bayesian variable selection procedure for generalized linear models. Specifically, we consider the consistency of a Bayes factor based on g-priors proposed by Sabanes Bove and Held (2011). The integrals necessary for the computation of this Bayes factor are performed with Laplace approximation and Gaussian quadrature. We show that, under certain regularity conditions, the resulting Bayes factor is consistent. Furthermore, a simulation study confirms our theoretical results. Finally, we illustrate this model selection procedure with an application to a real ecological dataset.

[1]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[2]  M. Gompper,et al.  Differential effects of experimental increases in sociality on ectoparasites of free-ranging raccoons. , 2010, The Journal of animal ecology.

[3]  M. Clyde,et al.  Mixtures of g Priors for Bayesian Variable Selection , 2008 .

[4]  Zehua Chen,et al.  EXTENDED BIC FOR SMALL-n-LARGE-P SPARSE GLM , 2012 .

[5]  Dean Phillips Foster,et al.  Calibration and empirical Bayes variable selection , 2000 .

[6]  Dani Gamerman,et al.  Sampling from the posterior distribution in generalized linear mixed models , 1997, Stat. Comput..

[7]  Guoqi Qian,et al.  Law of iterated logarithm and consistent model selection criterion in logistic regression , 2002 .

[8]  L. Wasserman,et al.  Computing Bayes Factors by Combining Simulation and Asymptotic Approximations , 1997 .

[9]  Leonhard Held,et al.  Hyper-$g$ priors for generalized linear models , 2010, 1008.1550.

[10]  Carlos Alberto de Bragança Pereira,et al.  Can a Significance Test Be Genuinely Bayesian , 2008 .

[11]  E. George.  The Variable Selection Problem , 2000 .

[12]  Richard J. Cook,et al.  Generalized Linear Model , 2014 .

[13]  Yuehua Wu,et al.  STRONG LIMIT THEOREMS ON MODEL SELECTION IN GENERALIZED LINEAR REGRESSION WITH BINOMIAL RESPONSES , 2006 .

[14]  D. Bates,et al.  Approximations to the Log-Likelihood Function in the Nonlinear Mixed-Effects Model , 1995 .

[15]  David W. Hosmer,et al.  Applied Logistic Regression , 1991 .

[16]  M. Gompper,et al.  The Effect and Relative Importance of Neutral Genetic Diversity for Predicting Parasitism Varies across Parasite Taxa , 2012, PloS one.

[17]  T. Ferguson.  A Course in Large Sample Theory , 1996 .

[18]  J. Naylor,et al.  Applications of a Method for the Efficient Computation of Posterior Distributions , 1982 .

[19]  J. E. Glynn,et al.  Numerical Recipes: The Art of Scientific Computing , 1989 .

[20]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

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

[22]  A. Zellner,et al.  Posterior odds ratios for selected regression hypotheses , 1980 .

[23]  M. Gompper,et al.  Biotic and abiotic predictors of tick (Dermacentor variabilis) abundance and engorgement on free-ranging raccoons (Procyon lotor) , 2007, Parasitology.

[24]  M. Steel,et al.  Benchmark Priors for Bayesian Model Averaging , 2001 .

[25]  H. Rue,et al.  Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations , 2009 .

[26]  S. Raudenbush,et al.  Maximum Likelihood for Generalized Linear Models with Nested Random Effects via High-Order, Multivariate Laplace Approximation , 2000 .

[27]  L. Tierney,et al.  Accurate Approximations for Posterior Moments and Marginal Densities , 1986 .

[28]  J. Ibrahim,et al.  Conjugate priors for generalized linear models , 2003 .

[29]  L. Fahrmeir,et al.  Correction: Consistency and Asymptotic Normality of the Maximum Likelihood Estimator in Generalized Linear Models , 1985 .