Group-linear empirical Bayes estimates for a heteroscedastic normal mean

DOI10.1080/01621459.2017.1280406zbMATH Open1398.62067arXiv1503.08503OpenAlexW779680562MaRDI QIDQ4962437FDOQ4962437

Authors: Asaf Weinstein, Zhuang Ma, Cun-Hui Zhang, Lawrence Brown 

Publication date: 2 November 2018

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: The problem of estimating the mean of a normal vector with known but unequal variances introduces substantial difficulties that impair the adequacy of traditional empirical Bayes estimators. By taking a different approach, that treats the known variances as part of the random observations, we restore symmetry and thus the effectiveness of such methods. We suggest a group-linear empirical Bayes estimator, which collects observations with similar variances and applies a spherically symmetric estimator to each group separately. The proposed estimator is motivated by a new oracle rule which is stronger than the best linear rule, and thus provides a more ambitious benchmark than that considered in previous literature. Our estimator asymptotically achieves the new oracle risk (under appropriate conditions) and at the same time is minimax. The group-linear estimator is particularly advantageous in situations where the true means and observed variances are empirically dependent. To demonstrate the merits of the proposed methods in real applications, we analyze the baseball data used in Brown (2008), where the group-linear methods achieved the prediction error of the best nonparametric estimates that have been applied to the dataset, and significantly lower error than other parametric and semi-parametric empirical Bayes estimators.

Full work available at URL: https://arxiv.org/abs/1503.08503

Recommendations

• Steinized empirical Bayes estimation for heteroscedastic data
• On general maximum likelihood empirical Bayes estimation of heteroscedastic IID normal means
• Double shrinkage empirical Bayesian estimation for unknown and unequal variances
• scientific article; zbMATH DE number 775917
• Estimations in the normal regression empirical bayes model

zbMATH Keywords

heteroscedasticityempirical Bayesasymptotic optimalityshrinkage estimatorcompound decision

Mathematics Subject Classification ID

Point estimation (62F10) Bayesian inference (62F15) Estimation in multivariate analysis (62H12) Ridge regression; shrinkage estimators (Lasso) (62J07) Analysis of variance and covariance (ANOVA) (62J10) Empirical decision procedures; empirical Bayes procedures (62C12)

Cites Work

• Adaptive wavelet estimation: A block thresholding and oracle inequality approach
• Title not available (Why is that?)
• General maximum likelihood empirical Bayes estimation of normal means
• Nonparametric empirical Bayes and compound decision approaches to estimation of a high-dimensional vector of normal means
• An empirical Bayes mixture method for effect size and false discovery rate estimation
• In-season prediction of batting averages: a field test of empirical Bayes and Bayes methodol\-ogies
• SURE Estimates for a Heteroscedastic Hierarchical Model
• Stein's Estimation Rule and Its Competitors--An Empirical Bayes Approach
• Estimation with Incompletely Specified Loss Functions (the Case of Several Location Parameters)
• Minimax estimators of the mean of a multivariate normal distribution
• A robust generalized Bayes estimator and confidence region for a multivariate normal mean
• Selecting a minimax estimator of a multivariate normal mean
• Admissible minimax estimation of a multivariate normal mean with arbitrary quadratic loss
• Title not available (Why is that?)
• The data-smoothing aspect of Stein estimates
• Improved minimax estimation of a multivariate normal mean under heteroscedasticity
• Steinized empirical Bayes estimation for heteroscedastic data

Cited In (23)

• A Regression Modeling Approach to Structured Shrinkage Estimation
• Nonparametric empirical Bayes improvement of shrinkage estimators with applications to time series
• Comment: Empirical Bayes, compound decisions and exchangeability
• An Empirical Bayes Method for Chi-Squared Data
• Empirical estimates for heteroscedastic hierarchical dynamic normal models
• Empirical Bayes Mean Estimation With Nonparametric Errors Via Order Statistic Regression on Replicated Data
• Heteroscedasticity-Adjusted Ranking and Thresholding for Large-Scale Multiple Testing
• On the nonparametric maximum likelihood estimator for Gaussian location mixture densities with application to Gaussian denoising
• Consistent and rate-optimal density estimation from heteroscedastic data groups
• Simultaneous estimation of normal means with side information
• Title not available (Why is that?)
• Poisson mean vector estimation with nonparametric maximum likelihood estimation and application to protein domain data
• Steinized empirical Bayes estimation for heteroscedastic data
• On general maximum likelihood empirical Bayes estimation of heteroscedastic IID normal means
• Adaptive sparse estimation with side information
• Title not available (Why is that?)
• Satisficing credibility for heterogeneous risks
• Optimal shrinkage estimation of predictive densities under \(\alpha\)-divergences
• Statistical theory powering data science
• Approximate Bayesian estimator for the parameter vector in linear models with multivariate t distribution errors
• Inference on a distribution from noisy draws
• A paradoxical argument about domination
• Confidence Intervals for Nonparametric Empirical Bayes Analysis

Uses Software

• mixfdr

This page was built for publication: Group-linear empirical Bayes estimates for a heteroscedastic normal mean

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4962437)