ON INVARIANCE AND MAXIMUM-LIKELIHOOD-ESTIMATION

被引：1

作者：

PAL, N

BERRY, JC

机构：

来源：

AMERICAN STATISTICIAN | 1992年 / 46卷 / 03期

关键词：

NUISANCE PARAMETER; ONE-TO-ONE FUNCTION; PARTIAL SUFFICIENCY; SPECIFIC SUFFICIENCY;

D O I：

10.2307/2685216

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：

The invariance of maximum likelihood estimators to non-one-to-one parameter transformations is considered. Three approaches to maximum likelihood estimation are discussed, and it is shown that invariance depends on the approach adopted. A theorem is provided that gives sufficient conditions for invariance to hold under all three approaches.

引用

页码：209 / 212

页数：4

共 6 条

[1] BARNDORFFNIELSO.OE, 1988, PARAMETRIC STATISTIC
[2] PARTIAL SUFFICIENCY - REVIEW
BASU, D
[J]. JOURNAL OF STATISTICAL PLANNING AND INFERENCE, 1978, 2 (01) : 1 - 13
[3] ELIMINATION OF NUISANCE PARAMETERS
BASU, D
[J]. JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1977, 72 (358) : 355 - 366
[4] Edwards A. F., 1972, LIKELIHOOD
[5] KALBFLEI.JD, 1970, J ROY STAT SOC B, V32, P175
[6] INVARIANCE OF MAXIMUM LIKELIHOOD ESTIMATORS
ZEHNA, PW
[J]. ANNALS OF MATHEMATICAL STATISTICS, 1966, 37 (03): : 744 - &

← 1 →