约束线性模型下回归系数的条件根方估计

本文提出了约束线性模型下回归系统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地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。

The Conditional Root Estimation of Regression Coefficient in Restricted Linear Model

In this paper, We proposed the conditional root estimator and the generalized root estimator of coefficient gin restricted linear model. We showed that they had smaller mean squares error than the RLSE,and discussed the admissibility of the two estimators and the choice of the unknoen parameters of the generalixed conditional root estimator.