Test for homogeneity in normal mixtures with unknown means and variances

被引:2
|
作者
Qin, Yongsong [1 ]
Smith, Bruce [2 ]
Lei, Qingzhu
机构
[1] Guangxi Normal Univ, Sch Math Sci, Guilin 541004, Guangxi, Peoples R China
[2] Dalhousie Univ, Dept Math & Stat, Halifax, NS, Canada
基金
中国国家自然科学基金;
关键词
Mixture model; Likelihood ratio test; Asymptotic distribution; LIKELIHOOD RATIO TEST; STRUCTURAL PARAMETER; MODELS;
D O I
10.1016/j.jspi.2009.06.002
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
This paper investigates the asymptotic properties of the likelihood ratio statistics for testing homogeneity in normal mixture models with unknown means and variances. The asymptotic null distribution of the ordinary likelihood ratio Statistic is chi(2)(2) under the conditions that the means are bounded, the variances are in a compact space away from 0 and infinity, and the mixture coefficients are away from 0. The asymptotic null distribution of a modified likelihood ratio statistic is also chi(2)(2) under the conditions that the means are bounded, and the variances are in a compact space away from 0 and infinity. (C) 2009 Elsevier B.V. All rights reserved.
引用
收藏
页码:4165 / 4178
页数:14
相关论文
共 50 条