【摘要】 考虑以下问题 :设n×m随机矩阵Y有分布N(Θn×m ,σ2 (Vn×n Σm×m) ) ,0 <σ2 ≤ 1 ,即Y服从均值向量为Θ协方差矩阵为σ2 (Vn×n Σm×m)的多元正态分布 ,其中 (Θ ,σ2 )为未知参数 .类似覃红讨论均值矩阵Θ的可估函数SΘ的线性估计AY在线性估计类中的泛容许性 .称Y的分布为矩阵正态分布更多还原

【Abstract】 If n×m random matrix Y is disted according to N(Θ n×m,σ 2(V n×nΣ m×m)),that is Y has a multivariate normal distribution with the mean vector Θ and the corariarce matrix σ 2 (V n×nΣ m×m),where Θ n×m and σ 2 are unknown,V and Σ are known nonnegative definite matrix.Let SΘ be linerly estimable.The necessary and sufficient condition for a liner estimator of SΘ to be general admissble among linear estimators under the matrix loss function be obtained.更多还原