正态均值向量参数的估计

Estimation of the Normal Mean Vector

本文给出了控制最小二乘估计的James-Stein型改进估计量,刻划了正态模型Y～N(Xβ,σ~2V),V可以奇异,σ~2∈S(?)(0,+∞),0(?)(?)下的线性可容许估计的特征。

Let Y～N.(β,σ~2I),β∈R^n, σ~2∈S(0,+∞), we obtain a 'James-Stein' form estimator, which dominates the LSE of β. For Y～N_n.(Xβ,σ~2V), V may be singular, β∈R^P,σ~2∈ S, 0, we give a characteristic of a linear admissible estimator.