摘要
本文对降秩多元线性模型的参数阵提出了一类有偏线性估计,讨论了许多重要的性质,从而把降秩模型中参数阵的估计问题转化为满秩模型中参数阵的估计问题.为了讨论方便,我们给出了一种特殊情况;最后,讨论了多元线性模型参数阵的Bayes线性估计.
In this paper we define a class of biased linear estimators for the parameter matrix in the multivariate linear model with deficient rank and discuss some important properties. We reduce the estimation problem in the deficient rank model to the one in the full-rank model. A special case of our class is given. Finally we discuss the Bayes linear estimator of parameter matrix in multivariate linear model.
出处
《湖南大学学报（自然科学版）》
EI
CAS
CSCD
1993年第6期6-11,24,共7页
Journal of Hunan University:Natural Sciences
关键词
多元线性模型
参数阵
有偏估计
multivariate linear model
least square estimator
mean square error
biased linear estimator
Bayes linear estimator
