ON A NECESSARY CONDITION FOR THE CONSISTENCY OF THE L1 ESTIMATES IN LINEAR-REGRESSION MODELS

被引：3

作者：

CHEN, XR

WU, Y

机构：

[1] CHINESE ACAD SCI,GRAD SCH,BEIJING,PEOPLES R CHINA

[2] YORK UNIV,DEPT MATH & STAT,N YORK M3J 1P3,ONTARIO,CANADA

来源：

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS | 1993年 / 22卷 / 03期

基金：

加拿大自然科学与工程研究理事会;

关键词：

LINEAR REGRESSION MODEL; L1; ESTIMATE; CONSISTENCY;

D O I：

10.1080/03610929308831043

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：

In the usual linear model Y(i) = x(i)'beta0 + e(i), i = 1, .... , n, denote by beta(n) the L1 estimate of beta0. Under some general conditions on {e(i)}, it is shown that SIGMA(i)infinity = \\x(i)\\ = infinity is a necessary condition for the consistency of beta(n).

引用

页码：631 / 639

页数：9