Hierarchical Bayes, maximum a posteriori estimators, and minimax concave penalized likelihood estimation

作者：Strawderman Robert L^*; Wells Martin T; Schifano Elizabeth D

来源：Electronic Journal of Statistics, 2013, 7(0): 973-990.

DOI：10.1214/13-EJS795

• Convex optimization
• Lasso penalty
• Moreau regularization
• minimax concave penalty
• sparsity
• smoothly clipped absolute deviation penalty
• thresholding

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%26apos;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.

• 出版日期2013