Maximum likelihood and restricted maximum likelihood estimators as functions of ordinary least squares and analysis of variance estimators

Autor:	Melinda H. McCann, Barry Kurt Moser
Rok vydání:	1996
Předmět:	Statistics and Probability Estimation theory Restricted maximum likelihood Ordinary least squares Statistics Statistics::Methodology Estimator Deviance (statistics) M-estimator Maximum likelihood sequence estimation Likelihood function Mathematics
Zdroj:	Communications in Statistics - Theory and Methods. 25:631-646
ISSN:	1532-415X 0361-0926
DOI:	10.1080/03610929608831718
Popis:	For a class of linear models with normally distributed error structures, necessary and sufficient conditions are given where the maximum likelihood (ML) and restricted maximum likelihood (REML) estimators of the parameters are functions of the ordinary least squares (OLS) and analysis of variance (ANOVA) estimators. These results are then used to extend the Gauss-Markov theorem to a broad class of error structures. The conditions are compared to equivalent results in the literature. Examples are presented to illustrate the problem.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::09600d5df1633620b883dfe139ba9918 https://doi.org/10.1080/03610929608831718 Zobrazit plný text záznamu