摘要
对于多元分析中的SUR回归系统,本文考虑系数βi的线性估计类,得到了该估计类中的最小协方差线性无偏估计^βi(),并应用Rao的协方差改进理论证明了^βi()是附加信息最优协方差改进估计。在其它备选估计所考虑的条件、或更弱的条件下,得到了非限定两步估计^βi(S)的有限样本方差结果。
In this paper,we consider an estimate class of βi in the system of Seemingly Unrelated Regression (SUR) equations.An optimal estimator Aβ^i() of βi is obtained within the class.An interpretation for this optimal estimator by Raos covariance adjustment theory is given,and some finite sample properties for the two-stage estimator Aβ^i(S) based on the unretricted estimate S of are also obtained under well known conditions.
出处
《五邑大学学报(自然科学版)》
CAS
1998年第3期9-15,共7页
Journal of Wuyi University(Natural Science Edition)
关键词
协方差改进估计
有限样本性质
SUR回归模型
seemingly unrelated regression equations
covariance adjustment estimate
two-stage estimate
finite sample variance matrix.