Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem

Abstract

This paper studies the multiplicity-correction effect of standard Bayesian variable-selection priors in linear regression. Our first goal is to clarify when, and how, multiplicity correction happens automatically in Bayesian analysis, and to distinguish this correction from the Bayesian Ockham's-razor effect. Our second goal is to contrast empirical-Bayes and fully Bayesian approaches to variable selection through examples, theoretical results and simulations. Considerable differences between the two approaches are found. In particular, we prove a theorem that characterizes a surprising aymptotic discrepancy between fully Bayes and empirical Bayes. This discrepancy arises from a different source than the failure to account for hyperparameter uncertainty in the empirical-Bayes estimate. Indeed, even at the extreme, when the empirical-Bayes estimate converges asymptotically to the true variable-inclusion probability, the potential for a serious difference remains.

Published in at http://dx.doi.org/10.1214/10-AOS792 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (2)

• A comparison of the Benjamini-Hochberg procedure with some Bayesian rules for multiple testing
• Nonparametric Bayesian multiple testing for longitudinal performance stratification

Cited by in corpus (26)

• Inferring causal impact using Bayesian structural time-series models
• Criteria for Bayesian model choice with application to variable selection
• Needles and Straw in a Haystack: Posterior concentration for possibly sparse sequences
• Overall Objective Priors
• DIAMONDS: a new Bayesian Nested Sampling tool. Application to Peak Bagging of solar-like oscillations
• The Horseshoe Estimator: Posterior Concentration around Nearly Black Vectors
• Metallicity effect on stellar granulation detected from oscillating red giants in open clusters
• Efficient inference for genetic association studies with multiple outcomes
• Computational approaches for empirical Bayes methods and Bayesian sensitivity analysis
• Nonparametric Bayesian multiple testing for longitudinal performance stratification
• Estimating causal structure using conditional DAG models
• Tight conditions for consistent variable selection in high dimensional nonparametric regression
• Inferring network structure from interventional time-course experiments
• Facial Keypoints Detection
• Causal network inference using biochemical kinetics
• Block Hyper-g Priors in Bayesian Regression
• Methods and Tools for Bayesian Variable Selection and Model Averaging in Univariate Linear Regression
• Bayes and empirical Bayes: do they merge?
• A Bayesian model averaging approach for observational gene expression studies
• Modular Bayes screening for high-dimensional predictors
• Towards a Multi-Subject Analysis of Neural Connectivity
• Laplace Approximation in High-dimensional Bayesian Regression
• Prioritizing covariates in the planning of future studies in the meta-analytic framework
• An approach for finding fully Bayesian optimal designs using normal-based approximations to loss functions
• Comment on Article by Berger, Bernardo, and Sun
• Handling Factors in Variable Selection Problems