Exponential Families and Variance Component Models
作者:
DennisL. Clason,
LeighW. Murray,
期刊:
The American Statistician
(Taylor Available online 1992)
卷期:
Volume 46,
issue 1
页码: 29-31
ISSN:0003-1305
年代: 1992
DOI:10.1080/00031305.1992.10475843
出版商: Taylor & Francis Group
关键词: Completeness;Lehmann-Scheffé theorem;Random components;Sufficiency
数据来源: Taylor
摘要:
This article shows that two slightly differential exponential families are needed for commonly used linear models. The first family is the usual textbook definition. The second family, due to a lemma by Gautschi, is needed for some cross-classified variance component models.
点击下载:
PDF (295KB)
返 回