摘要
本文在矩阵损失函数下。
In this paper, we consider the problem of the minimax admissibility characterization on linear eslinates of multivariate regression coefficient for grouth curver model under matrix loss. The necessary and sufficlent conditions for a linear estimator DYF(DYF+H) to be minimax admissible among the class of homogeneous (nonhomegeneous) linear estimators are given.
出处
《数学杂志》
CSCD
1997年第2期163-168,共6页
Journal of Mathematics
基金
国家自然科学基金
