A variant of AIC based on the Bayesian marginal likelihood

DOI10.1007/S13571-018-0152-7zbMATH Open1392.62206arXiv1503.07102OpenAlexW2964252496MaRDI QIDQ721607FDOQ721607

Yuki Kawakubo, Muni S. Srivastava, Tatsuya Kubokawa

Publication date: 19 July 2018

Published in: Sankhyā. Series B (Search for Journal in Brave)

Abstract: We propose information criteria that measure the prediction risk of a predictive density based on the Bayesian marginal likelihood from a frequentist point of view. We derive criteria for selecting variables in linear regression models, assuming a prior distribution of the regression coefficients. Then, we discuss the relationship between the proposed criteria and related criteria. There are three advantages of our method. First, this is a compromise between the frequentist and Bayesian standpoints because it evaluates the frequentist's risk of the Bayesian model. Thus, it is less influenced by a prior misspecification. Second, the criteria exhibits consistency when selecting the true model. Third, when a uniform prior is assumed for the regression coefficients, the resulting criterion is equivalent to the residual information criterion (RIC) of Shi and Tsai (2002).

Full work available at URL: https://arxiv.org/abs/1503.07102

Recommendations

• scientific article
• ON VARIABLE SELECTION IN LINEAR REGRESSION
• The risk inflation criterion for multiple regression
• Empirical Bayes vs. fully Bayes variable selection
• Regression Model Selection—A Residual Likelihood Approach

zbMATH Keywords

variable selectionBICconsistencyAIClinear regression modelKullback-Leibler divergenceresidual information criterion

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Linear regression; mixed models (62J05)

Cites Work

• Estimating the dimension of a model
• Regression and time series model selection in small samples
• Bayesian Measures of Model Complexity and Fit
• Title not available (Why is that?)
• Title not available (Why is that?)
• A new look at the statistical model identification
• Recovery of inter-block information when block sizes are unequal
• A review of Bayesian variable selection methods: what, how and which
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Asymptotic properties of criteria for selection of variables in multiple regression
• Model selection and Akaike's information criterion (AIC): The general theory and its analytical extensions
• On Bayesian model and variable selection using MCMC
• Conditional Akaike information for mixed-effects models
• Variable selection for regression models
• Further analysis of the data by Akaike's information criterion and the finite corrections
• Bayesian predictive information criterion for the evaluation of hierarchical Bayesian and empirical Bayes models
• Conditional information criteria for selecting variables in linear mixed models
• Bayesian variable selection and estimation for group Lasso
• An optimal selection of regression variables
• Regression Model Selection—A Residual Likelihood Approach
• On the use of the predictive likelihood of a Gaussian model
• Information criteria for the predictive evaluation of bayesian models

Cited In (3)

• ON VARIABLE SELECTION IN LINEAR REGRESSION
• Consistent Bayesian information criterion based on a mixture prior for possibly high‐dimensional multivariate linear regression models
• Bayesian prediction and model selection for locally asymptotically mixed normal models

This page was built for publication: A variant of AIC based on the Bayesian marginal likelihood

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q721607)