【摘要】 文中给出了多元线性模型 ( Y,XΘ,V Σ)中回归系数的线性估计在线性估计类中是泛可容许的一些特征 ,其中损失函数取为 ( D( Y) - SΘ )′( D( Y) - SΘ) ,参数Θ和Σ取值于约束集合HN={( Θ,Σ) :Θ′X′NXΘ≤Σ,N≥ 0 }更多还原

【Abstract】 In this paper, we will give some characters of the universal admissibility for linear estimators of regression coefficients under the multivariate linear models (Y,XΘ,VΣ) and the matrix loss function (D(Y)-SΘ)′(D(Y)-SΘ), where considered parameters Θ and Σ vary in the restricted class H\-N={(Θ,Σ):Θ′X′NXΘ≤Σ,N≥0}. The shown results shall establish the relationship between the linear admissible estimators of SΘ under the multivariate linear model (Y,XΘ,VΣ) and the linear admissible estimators of Sβ under the Gauss Markov model (Y,Xβ,σ\+2V) and extend some results in literature.更多还原