矩阵损失下带约束的回归系数非齐次线性估计的Minimax可容许性的特征  

The Minimax Admissibility Characterization of Nonhomogeneous Linear Estimates with Respect to Restricted Regression Coefficients under Matrix Loss Function

作　　者：谭萄[1] 王成名[2] 王炜炘[2] 

机构地区：[1]天津财经学院统计系,天津300222 [2]广西师范大学数学与计算机科学学院,桂林541004

出　　处：《应用概率统计》2002年第1期1-7,共7页Chinese Journal of Applied Probability and Statistics

摘　　要：本文在矩阵损失下给出了带约束的回归系数线性估计在非齐次线性估计类中是Ｍｉｎｉｍａｘ可容许估计的充要条件．In this paper, we consider the Minimax admissibility characterization of linear estimates with respect to restricted regression coefficient under a matrix loss function. The necessary and sufficient conditions are given for a linear estimates AY+c to be Minimax admissible in the class of nonhomogeneous linear estimates.

关 键 词：矩阵损失 MINIMAX可容许估计 非齐次线性估计 回归系数 

分 类 号：O211.67[理学—概率论与数理统计] O212.1[理学—数学]