On predictive density estimation for location families under integrated squared error loss

DOI10.1016/J.JMVA.2015.07.013MaRDI QIDQ893168zbMATH OpenOpenAlexFDO

Authors Tatsuya Kubokawa, Éric Marchand, William E. Strawderman

Publication date 13 November 2015

Published in Journal of Multivariate Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1408.5297

zbMATH Keywords

minimaxconvolutionsmultivariate normaldominanceinadmissibilityloss functionpredictive densityrestricted parameter spaceintegrated squared errorStein estimationfrequentist riskminimum risk equivariantscale mixture of normalsconcave loss

Mathematics Subject Classification ID

Point estimation (62F10) Bayesian inference (62F15) Minimax procedures in statistical decision theory (62C20) Parametric inference under constraints (62F30) Statistical decision theory and fuzziness (62C86)

Abstract: Our investigation concerns the estimation of predictive densities and a study of efficiency as measured by the frequentist risk of such predictive densities with integrated $L_2$ and $L_1$ losses. Our findings relate to a $p-$variate spherically symmetric observable $Xsim{p}_X(|x-mu{|}^2)$ and the objective of estimating the density of $Ysim{q}_Y(|y-mu{|}^2)$ based on $X$. For $L_2$ loss, we describe Bayes estimation, minimum risk equivariant estimation (MRE), and minimax estimation. We focus on the risk performance of the benchmark minimum risk equivariant estimator, plug-in estimators, and plug-in type estimators with expanded scale. For the multivariate normal case, we make use of a duality result with a point estimation problem bringing into play reflected normal loss. In three of more dimensions (i.e., $pgeq3$), we show that the MRE estimator is inadmissible under $L_2$ loss and provide dominating estimators. This brings into play Stein-type results for estimating a multivariate normal mean with a loss which is a concave and increasing function of $|hatmu-mu{|}^2$. We also study the phenomenon of improvement on the plug-in density estimator of the form $q_Y(|y-aX{|}^2),,0<aleq1,,$ by a subclass of scale expansions $frac1c^p,q_Y(|(y-aX)/c{|}^2)$ with $c>1$, showing in some cases, inevitably for large enough $p$, that all choices $c>1$ are dominating estimators. Extensions are obtained for scale mixture of normals including a general inadmissibility result of the MRE estimator for $pgeq3$. Finally, we describe and expand on analogous plug-in dominance results for spherically symmetric distributions with $pgeq4$ under $L_1$ loss.

Recommendations

• On predictive density estimation for location families under integrated absolute error loss
• Inadmissibility of the best equivariant predictive density in the unknown variance case
• On efficient prediction and predictive density estimation for normal and spherically symmetric models
• Minimaxity and admissibility of predictive density estimators under S-Hellinger distances
• On the Stein effect under density power divergence loss

Cites work

• scientific article; zbMATH DE number 51418 (Why is no real title available?)
• scientific article; zbMATH DE number 1220667 (Why is no real title available?)
• scientific article; zbMATH DE number 1300859 (Why is no real title available?)
• scientific article; zbMATH DE number 3442988 (Why is no real title available?)
• scientific article; zbMATH DE number 3070795 (Why is no real title available?)
• A Family of Minimax Estimators of the Mean of a Multivariate Normal Distribution
• A Note on the Estimation of Probability Density Functions
• A shrinkage predictive distribution for multivariate normal observables
• Admissible predictive density estimation
• Bayesian improvements of a MRE estimator of a bounded location parameter
• Estimation in restricted parameter spaces: a review
• Estimation of the mean of a multivariate normal distribution
• Exact Minimax Strategies for Predictive Density Estimation, Data Compression, and Model Selection
• Frequentist prediction intervals and predictive distributions
• Goodness of prediction fit
• Improved minimax predictive densities under Kullback-Leibler loss
• Improving on the minimum risk equivariant estimator of a location parameter which is constrained to an interval or a half-interval
• Minimax estimation of location parameters for certain spherically symmetric distributions
• Minimax estimation of location parameters for spherically symmetric distributions with concave loss
• Minimaxity in predictive density estimation with parametric constraints
• On improved predictive density estimation with parametric constraints
• On improved shrinkage estimators for concave loss
• On the construction of Bayes minimax estimators
• On the density of the sum of two independent Student \(t\)-random vectors
• On the sum of \(t\) and Gaussian random variables
• Shrinkage estimators of the location parameter for certain spherically symmetric distributions
• The reflected normal loss function
• Uniform priors on convex sets improve risk
• Variation Diminishing Transformations: A Direct Approach to Total Positivity and its Statistical Applications

Cited in

(14)

• Inadmissibility of the best equivariant predictive density in the unknown variance case
• Predictive density estimators with integrated \(L_1\) loss
• Novel credibility approaches for Lorenz curve and Gini coefficient estimation
• On the Stein effect under density power divergence loss
• On predictive density estimation for location families under integrated absolute error loss
• Bayesian inference and prediction for mean-mixtures of normal distributions
• Predictive density estimation under the Wasserstein loss
• On minimax optimality of sparse Bayes predictive density estimates
• Bayesian estimation and prediction for certain type of mixtures
• Minimaxity and admissibility of predictive density estimators under S-Hellinger distances
• On predictive density estimation with additional information
• On predictive density estimation for gamma models with parametric constraints
• On predictive density estimation under \(\alpha\)-divergence loss
• On efficient prediction and predictive density estimation for normal and spherically symmetric models

This page was built for publication: On predictive density estimation for location families under integrated squared error loss

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q893168)