In combination with the Bayes estimator for the parameter of linear exponential model under the same loss function, the nonparametric empirical Bayes estimator of the unknown parameter was obtained.

Abstract: Under the matrix loss function, the necessary and sufficient conditions of linear admissible estimates of nonestimatible parameter functions for a general linear model are obtained.

We investigate the admissibility of the linear estimate of random regression coefficients under a matrix loss function in general growth curve models.