Open Access
2013 Hierarchical Bayes, maximum a posteriori estimators, and minimax concave penalized likelihood estimation
Robert L. Strawderman, Martin T. Wells, Elizabeth D. Schifano
Electron. J. Statist. 7: 973-990 (2013). DOI: 10.1214/13-EJS795

Abstract

Priors constructed from scale mixtures of normal distributions have long played an important role in decision theory and shrinkage estimation. This paper demonstrates equivalence between the maximum aposteriori estimator constructed under one such prior and Zhang’s minimax concave penalization estimator. This equivalence and related multivariate generalizations stem directly from an intriguing representation of the minimax concave penalty function as the Moreau envelope of a simple convex function. Maximum aposteriori estimation under the corresponding marginal prior distribution, a generalization of the quasi-Cauchy distribution proposed by Johnstone and Silverman, leads to thresholding estimators having excellent frequentist risk properties.

Citation

Download Citation

Robert L. Strawderman. Martin T. Wells. Elizabeth D. Schifano. "Hierarchical Bayes, maximum a posteriori estimators, and minimax concave penalized likelihood estimation." Electron. J. Statist. 7 973 - 990, 2013. https://doi.org/10.1214/13-EJS795

Information

Published: 2013
First available in Project Euclid: 15 April 2013

zbMATH: 1337.62172
MathSciNet: MR3044506
Digital Object Identifier: 10.1214/13-EJS795

Subjects:
Primary: 62C60 , 62J07

Keywords: Convex optimization , lasso penalty , minimax concave penalty , Moreau regularization , smoothly clipped absolute deviation penalty , Sparsity , thresholding

Rights: Copyright © 2013 The Institute of Mathematical Statistics and the Bernoulli Society

Back to Top