A COMPARISON OF MAXIMUM-LIKELIHOOD AND QUASI-MINIMAX APPROACH TO MISSING VALUES

被引:1
作者
JANNER, M [1 ]
STAHLECKER, P [1 ]
机构
[1] UNIV OLDENBURG,W-2900 OLDENBURG,GERMANY
关键词
LINEAR REGRESSION MODEL; MISSING VALUES; MAXIMUM LIKELIHOOD ESTIMATION; PRIOR INFORMATION; MINIMAX ESTIMATION;
D O I
10.1080/03610929308831022
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We consider the problem of estimating the parameter vector in the linear model when the observations on the independent variables are partially missing. The new quasi minimax approach, which uses prior restrictions on the exogenous variables, is compared to the Maximum Likelihood method with respect to the (empirical) mean square error.
引用
收藏
页码:319 / 333
页数:15
相关论文
共 14 条