Consistency of Bayes factor for nonnested model selection when the model dimension grows

DOI10.3150/15-BEJ720MaRDI QIDQ726729zbMATH OpenWikidataFDO

Authors Min Wang, Yuzo Maruyama

Publication date 14 July 2016

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1503.06155, https://projecteuclid.org/euclid.bj/1462297675

zbMATH Keywords

Bayes factorgrowing number of parametersmodel selection consistencynonnested linear models

Mathematics Subject Classification ID

Bayesian inference (62F15) Linear regression; mixed models (62J05)

Abstract: Zellner's $g$-prior is a popular prior choice for the model selection problems in the context of normal regression models. Wang and Sun [J. Statist. Plann. Inference 147 (2014) 95-105] recently adopt this prior and put a special hyper-prior for $g$, which results in a closed-form expression of Bayes factor for nested linear model comparisons. They have shown that under very general conditions, the Bayes factor is consistent when two competing models are of order $O(n^{au})$ for $au<1$ and for $au=1$ is almost consistent except a small inconsistency region around the null hypothesis. In this paper, we study Bayes factor consistency for nonnested linear models with a growing number of parameters. Some of the proposed results generalize the ones of the Bayes factor for the case of nested linear models. Specifically, we compare the asymptotic behaviors between the proposed Bayes factor and the intrinsic Bayes factor in the literature.

Review text

The paper under review investigates Bayes factor consistency associated with Zellner's g-prior for the problem of comparing nonnested linear models with a growing number of parameters under three selected asymptotic scenarios of the model dimension growth with the sample size. Specifically, the authors compare the asymptotic results between the proposed Bayes factor and the intrinsic Bayes factor due to \textit{F. J. Girón} et al. [Rev. R. Acad. Cienc. Exactas Fís. Nat., Ser. A Mat., RACSAM 104, No. 1, 57--67 (2010; Zbl 1193.62036)]. Some of the obtained results generalize the ones of the Bayes factor for the case of nested linear models in [\textit{M. Wang} and \textit{X. Sun}, J. Stat. Plann. Inference 147, 95--105 (2014; Zbl 1432.62216)].

Recommendations

• Bayes factor consistency for nested linear models with a growing number of parameters
• Consistency of objective Bayes factors for nonnested linear models and increasing model dimension
• Consistency of objective Bayes factors as the model dimension grows
• Consistency of Bayes factors under hyper g-priors with growing model size
• Posterior consistency of \(g\)-prior for variable selection with a growing number of parameters

Cites work

• Approximations and consistency of Bayes factors as model dimension grows
• Bayes factor consistency for nested linear models with a growing number of parameters
• Bayes factor consistency for unbalanced ANOVA models
• Bayes Factors
• Bayesian structured variable selection in linear regression models
• Benchmark priors for Bayesian model averaging.
• Comparison of Bayesian objective procedures for variable selection in linear regression
• Conservative prior distributions for variance parameters in hierarchical models
• Consistency of Bayesian procedures for variable selection
• Consistency of objective Bayes factors as the model dimension grows
• Consistency of objective Bayes factors for nonnested linear models and increasing model dimension
• Criteria for Bayesian model choice with application to variable selection
• Fully Bayes factors with a generalized g-prior
• scientific article; zbMATH DE number 4070082 (Why is no real title available?)
• scientific article; zbMATH DE number 409719 (Why is no real title available?)
• scientific article; zbMATH DE number 3213229 (Why is no real title available?)
• Mixtures of g Priors for Bayesian Variable Selection
• Mixtures of \(g\)-priors for Bayesian model averaging with economic applications
• NONNESTED LINEAR MODEL SELECTION REVISITED
• Objective Testing Procedures in Linear Models: Calibration of the p‐values
• On the Choice of Forecasting Formulas
• Posterior model consistency in variable selection as the model dimension grows
• The behaviour of linear model selection tests under globally non-nested hypotheses
• Two-way ANOVA models with unbalanced data

Cited in

(11)

• Posterior consistency of \(g\)-prior for variable selection with a growing number of parameters
• Approximations and consistency of Bayes factors as model dimension grows
• Bayesian and frequentist evidence in one-sided hypothesis testing
• Bayes factor asymptotics for variable selection in the Gaussian process framework
• Bayesian bandwidth test and selection for high-dimensional banded precision matrices
• Consistency of Bayes factors under hyper g-priors with growing model size
• Bayes factor consistency for nested linear models with a growing number of parameters
• Efficient functional ANOVA through wavelet-domain Markov groves
• Bayesian analysis of testing general hypotheses in linear models with spherically symmetric errors
• Consistency of objective Bayes factors for nonnested linear models and increasing model dimension
• Consistency of objective Bayes factors as the model dimension grows

This page was built for publication: Consistency of Bayes factor for nonnested model selection when the model dimension grows

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q726729)