半相依回归系统在一类约束条件下的协方差改进估计  

Covariance-improved Estimator of Seemingly Unrelated Regression Equation System under a Kind of Restraint Condition

作　　者：阳宁光[1] 

机构地区：[1]广东商学院数学部,广东广州510320

出　　处：《广西师范学院学报（自然科学版）》2005年第1期27-30,34,共5页Journal of Guangxi Teachers Education University(Natural Science Edition)

摘　　要：对于由两个误差项相关的线性回归方程组成的SUR系统,通常地根据以往的知识经验可以获得有关未知参数β的一些信息,该文就是在线性约束r=Rβ下,讨论SUR系统的协方差改进估计序列及两步协方差改进估计序列的收敛性问题.For a Seemingly Unrelated Regression with tow linear regression models, we can usually obtain some information about unknown parameter according to past knowledge experience.In this paper we try to obtain the covariance-improved estimator sequence and the convergence of two-stage covariance-improved estimator sequence.

关 键 词：线性约束 协方差改进估计 两步估计 MDE准则 

分 类 号：O212.2[理学—概率论与数理统计]