All \(E\)-admissible and \(G\)-admissible linear estimators of the common mean matrix parameter. (Chinese. English summary) Zbl 0936.62062

Peng, Junhao; Qin, Hong; Wu, Min

J. Cent. China Norm. Univ., Nat. Sci. 32, No. 4, 400-404 (1998).

Summary: For a general multivariate linear model, this paper discusses the \(E\)-admissibility and \(G\)-admissibility of the linear estimator for \(SX \Theta\) in the class of all estimators, and gives a necessary and sufficient condition for \(\sum^n_{i=1}A_iY_i\) to be an \(E\)-admissible estimator of \(SX \Theta\) and that for \(\sum^m_{i=1}A_iY_i\) to be a \(G\)-admissible estimator of \(SX \Theta\).

MSC:

62H12	Estimation in multivariate analysis
62C15	Admissibility in statistical decision theory
62J05	Linear regression; mixed models

Keywords:

common mean matrix parameter; \(E\)-admissibility; \(G\)-admissibility