多元线性模型中最小二乘估计新的相对效率  

A New Relative Efficiency of the Least Squares Estimator in Multivariate Linear Model

作　　者：侯景臣[1] 

机构地区：[1]辽宁石油化工大学,辽宁抚顺113001

出　　处：《科技通报》2004年第5期385-387,391,共4页Bulletin of Science and Technology

摘　　要：定义了广义行列式的概念,并在多元线性模型中,引入了回归系数的最小二乘估计与最佳线性无估计的一种新的相对效率.利用广义行列式的性质和矩阵分析的方法,通过对相对效率下界的研究,推广了Bloomfield Watson定理.A kind of generalized determinant is defined and a new relative efficiency are introduced by the least square estimator of regression coefficient comparing to the best linear unbiased estimator in the multivariate linear model. Using the propositions of generality determinant and matrix analysis,and researching the lower bound of relative efficiency,the Bloomfield-Watson's theorem has been extended.

关 键 词：多元分析 最小二乘估计 相对效率 Bloomfield—Watson不等式 

分 类 号：O212.1[理学—概率论与数理统计]