正态线性试验中可估函数的最小最大估计On Minimax Estimators of Estimable Funetions in Normal Linear Experiments
姜峰
摘要(Abstract):
对于正态线性试验NL(Xβ,δ~2V),V为已知κ×n阶正定矩阵,δ~2为未知正参数,通过容许性理论,在平方损失函数(δ~2+β~rX~rV~(-1)Xβ)~(-1)‖δ-SXβ‖下,本文证明了SXβ的线性估计是所有估计类中一致最小最大估计。
关键词(KeyWords): 正态线性试验;平方损失;可估线性函数;最小最大估计;容许性理论
基金项目(Foundation):
作者(Author): 姜峰
DOI: 10.16393/j.cnki.37-1436/z.1999.04.003
参考文献(References):
- 1 khursheed A, A family of aadmissable minimax estimator of the mean of a multivariate normal distribution, Ann. Statist., 1979,(1) :517~525
- 2 Baranchik,Λ.J_i,Λ. Family of minimax cstimator of the mcan of a. multivariate normal distribution, Ann. Math. Statist, 1970,(11) :642~645
- 3 Efron B. Monts C., Familles of minimax estimators of the mean of a multivariate normad distribution, Ann. Statist, 1976, (4) :11~21
- 4 徐兴忠.二次损失下回归系数的线性估计.数学年刊,14A,993,(5) :621~628
- 5 陈希孺,等.线性模型参数的估计理论.北京:科学出版社,1985