综合c-K条件岭估计及其优良性  

The integrated c- K conditional ridge estimation and superiority

作　　者：王志福[1] 刘佳瑞[1] 

机构地区：[1]渤海大学数理学院,辽宁锦州121013

出　　处：《渤海大学学报（自然科学版）》2016年第2期97-100,共4页Journal of Bohai University:Natural Science Edition

基　　金：国家自然科学基金项目(No:11671070)

摘　　要：对齐次等式约束的线性回归模型回归系数提出一个新的有偏估计-综合c-K条件岭估计.研究综合c-K条件岭估计的性质,并证明存在一定条件,使其在均方误差意义下优于约束最小二乘估计、条件岭估计、广义条件岭估计以及综合c-K条件岭估计.A new biased estimation of regression coefficient in a regression model with homogeneous equality constraints,the integrated c- K conditional ridge estimation,is proposed. The properties of the Integrated c- K conditional ridge estimation are studied,and under some conditions,it is proved that it＇s better than the constrained least square estimation,the conditional ridge estimation,the generalized conditional ridge estimation and the integrated c- K conditional ridge estimation in the sense of the mean square error.

关 键 词：综合c-K条件岭估计 线性回归模型 约束最小二乘估计 均方误差 

分 类 号：O212.2[理学—概率论与数理统计]