Estimation under matrix quadratic loss and matrix superharmonicity
- 20 April 2021
- journal article
- research article
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 109 (2) , 503-519
Abstract
We investigate estimation of a normal mean matrix under the matrix quadratic loss. Improved estimation under the matrix quadratic loss implies improved estimation of any linear combination of the columns under the quadratic loss. First, an unbiased estimate of risk is derived and the Efron–Morris estimator is shown to be minimax. Next, a notion of matrix superharmonicity for matrix-variate functions is introduced and shown to have analogous properties with usual superharmonic functions, which may be of independent interest. Then, we show that the generalized Bayes estimator with respect to a matrix superharmonic prior is minimax. We also provide a class of matrix superharmonic priors that includes the previously proposed generalization of Stein’s prior. Numerical results demonstrate that matrix superharmonic priors work well for low rank matrices.Keywords
This publication has 15 references indexed in Scilit:
- Singular value shrinkage priors for Bayesian predictionBiometrika, 2015
- Potential Theory2014
- Shrinkage estimation with a matrix loss functionElectronic Journal of Statistics, 2012
- Shrinkage priors for Bayesian predictionThe Annals of Statistics, 2006
- Improved minimax predictive densities under Kullback–Leibler lossThe Annals of Statistics, 2006
- All Admissible Linear Estimates of the Mean MatrixJournal of Multivariate Analysis, 1993
- Minimax estimators in the manova model for arbitrary quadratic loss and unknown covariance matrixJournal of Multivariate Analysis, 1991
- Minimax estimators in the normal MANOVA modelJournal of Multivariate Analysis, 1989
- Minimax Bayes estimators of a multivariate normal meanJournal of Multivariate Analysis, 1978
- Empirical Bayes on vector observations: An extension of Stein's methodBiometrika, 1972