Covariance matrix estimation under data-based loss

DOI10.1016/J.SPL.2021.109160MaRDI QIDQ2244574zbMATH OpenOpenAlexFDO

Authors Dominique Fourdrinier, Anis M. Haddouche, Fatiha Mezoued

Publication date 12 November 2021

Published in Statistics \& Probability Letters (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2012.11920

zbMATH Keywords

data-based losselliptically symmetric distributionshigh-dimensional statisticsorthogonally invariant estimatorsStein-Haff type-identities

Mathematics Subject Classification ID

Statistical decision theory (62C99) Point estimation (62F10) Estimation in multivariate analysis (62H12)

Abstract: In this paper, we consider the problem of estimating the $pimesp$ scale matrix $Sigma$ of a multivariate linear regression model $$ when the distribution of the observed matrix $Y$ belongs to a large class of elliptically symmetric distributions. After deriving the canonical form $(Z^T{U}^T{)}^T$ of this model, any estimator $hatSigma$ of $Sigma$ is assessed through the data-based loss tr$(S^{+}Sigma,(Sigm{a}^{-1}hatSigma-I_p{)}^2),$ where $S=U^{T}U$ is the sample covariance matrix and $S^{+}$ is its Moore-Penrose inverse. We provide alternative estimators to the usual estimators $a,S$, where $a$ is a positive constant, which present smaller associated risk. Compared to the usual quadratic loss tr$(Sigm{a}^{-1}hatSigma-I_p{)}^2$, we obtain a larger class of estimators and a wider class of elliptical distributions for which such an improvement occurs. A numerical study illustrates the theory.

Recommendations

• Scale matrix estimation of an elliptically symmetric distribution in high and low dimensions
• scientific article; zbMATH DE number 977893
• Estimation of scale matrix of elliptically contoured matrix distributions
• Estimating the covariance matrix and the generalized variance under a symmetric loss
• Robust empirical Bayes estimation of the elliptically countoured covariance matrix

Cites work

• A unified approach to estimating a normal mean matrix in high and low dimensions
• An orthogonally invariant minimax estimator of the covariance matrix of a multivariate normal population
• Empirical Bayes estimation of the multivariate normal covariance matrix
• Estimation of a high-dimensional covariance matrix with the Stein loss
• Estimation of the precision matrix of a singular Wishart distribution and its application in high-dimensional data
• Estimation with quadratic loss.
• scientific article; zbMATH DE number 48335 (Why is no real title available?)
• Improved second order estimation in the singular multivariate normal model
• Minimax estimation of location vectors for a wide class of densities
• Multivariate empirical Bayes and estimation of covariance matrices
• On Wishart distribution: some extensions
• Robust improvement in estimation of a covariance matrix in an elliptically contoured distribution
• Robust improvement in estimation of a mean matrix in an elliptically contoured distribution
• Robust minimax Stein estimation under invariant data-based loss for spherically and elliptically symmetric distributions
• Scale matrix estimation of an elliptically symmetric distribution in high and low dimensions
• Shrinkage estimation
• Shrinkage estimation for mean and covariance matrices
• Shrinkage estimators for large covariance matrices in multivariate real and complex normal distributions under an invariant quadratic loss
• Singular Wishart and multivariate beta distributions
• Unbiased risk estimates for matrix estimation in the elliptical case
• Unbiased Risk Estimates for Singular Value Thresholding and Spectral Estimators
• Unified improvements in estimation of a normal covariance matrix in high and low dimensions

Cited in

(6)

• Improved estimation of a covariance matrix under quadratic loss
• A unified approach for covariance matrix estimation under Stein loss
• Scale matrix estimation of an elliptically symmetric distribution in high and low dimensions
• scientific article; zbMATH DE number 4169892 (Why is no real title available?)
• Truncated Estimators for a Precision Matrix
• Noise covariance estimation in multi-task high-dimensional linear models

This page was built for publication: Covariance matrix estimation under data-based loss

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2244574)