摘要
本文提出了一类相依回归系统参数的广义两步有偏估计,证明了当设计矩阵呈病态时,广义两步有偏估计的均方误差小于两步协方差改进估计的均方误差.
We put forward a generalized two steps biased estimate and prove that the mean square error of the generalized two steps biased estimate is less than the mean square error of the two steps covariance improvement estimate when the designed matrix is ill-condition.
出处
《漳州师范学院学报（自然科学版）》
2005年第2期28-31,共4页
Journal of ZhangZhou Teachers College（Natural Science)
基金
漳州师范学院科研基金资助项目(SK04008)
福建省教育厅科研基金资助项目(JA04248)
关键词
有偏估计
系统参数
相依回归
广义
协方差改进估计
均方误差
设计矩阵
seemingly unrelated regression system
two steps covariance improvement estimate
mean square error
generalized two steps biased estimate
