ADMISSIBILITY OF LINEAR ESTIMATORS OF THE COMMON MEAN VECTOR IN LINEAR MODEL
Eifang Zhejiang · 1991
Consider the linear model: Y_i=(y_1~(i),…,y_n~(i))'=β+e_i=(β_1,…, β_n)＇+(e_1~(i)…,e_n~(i))＇ i=1,…,m where Y_1,…, Y_m are independent, E (e_i)=0, Var (e_i)=σ_i~2V,i= 1,…, m. β∈R~n, 0σ_i~2 ∞ unknown. The necessary and sufficient conditions that sum from i=1 to m A_iY_i and sum from i=1 to m A_iY_i+C are admissible for Sβ within the classes {sum from i=1 to m A_iY_i; A_i: s×n} and {sum from i=1 to m A_iY_i+C; A_i: s×n, C: s×1} are given respectively. In comparision with the method used in other papers, we use the method of matrix derivative.