摘要
提出了约束线性模型下回归系数的条件根方估计和广义条件根方估计，证明了在一定的条件下，两者在均方误差意义都能很好改进回归系数的约束最小二乘估计，并讨论了它们的可容许性及广义条件根方估计中未知参数的选取方法．
In this paper, we propose the conditional root root estimt and the generlize root root estimat of coefficient β in restricted linear model. We showe that they have smaller mean squares error than the RLSE, and discuss the admissibility of the two estimatrs and the choice of the unknown parmeters of the generalized conditional root root estimator.
出处
《沈阳师范学院学报（自然科学版）》
1999年第4期13-19,共7页
Journal of Shenyang Normal University(Natural Science)
关键词
条件根方估计
可容许性
约束线性模型
回归系数
Mean squares error
Conditional root root estimator
Generalized root root estimator
Admissibility
