摘要
本文证明了一个关于SUR模型回归系数最小方差线性无偏估计(MVLUE)的充要条件,并利用此充要条件讨论了几类SUR模型回归系数的MVLUE估计及两步估计.在方法上避免了与分块矩阵求逆有关的运算,所得结论推广和完善了已有的一些结果.
This paper proves a necessary and sufficient condition about the minimum variance linear unbiased estimators for the regression coefficients of the SUR model. Several classes of the MVLUE and the two stage estimators for the regression coefficients of the SUR model are discussed by means of this result. This method avoids some calculations about the inverse of the block matrices. The paper also extends and perfects some results given by other authors.
出处
《应用数学》
CSCD
北大核心
1991年第1期83-89,共7页
Mathematica Applicata
关键词
SUR模型
回归系数
估计
Seemingly unrelated regression model
Regression coefficient
Minimum variance linear unbiased estimator
Two stage estimators
