|
|
Registro Completo |
Biblioteca(s): |
Embrapa Agricultura Digital. |
Data corrente: |
26/03/1998 |
Data da última atualização: |
14/12/2010 |
Autoria: |
GILCHRIST, R.; PORTIDES, G. |
Título: |
Using GLIM4 to estimate the tuning constant for Huber's M-estimate of location. |
Ano de publicação: |
1994 |
Fonte/Imprenta: |
In: DUTTER, R.; GROSSMANN, W. (Ed.). Compstat: proceedings in computational statistics. [Hildeberg]: Physica-Verlag, 1994. |
Páginas: |
p. 251-256. |
ISBN: |
3-7908-0793-1 |
Idioma: |
Inglês |
Conteúdo: |
Probably the best well known technique for the resistant estimation of a location parameter is Huber's (1964) proposal for an M-estimate of location for the e-contaminated Normal distribution. However, a disadvantage of this technique is the assumption that the amount of contamination, and hence the so called tuning constant is known, a priori. In the present paper the authors present huber's proposal and how it can be implemented using the iteratively te-weighted least dquares (IRLS) procedure used in GLIM4. A proposal for estimating the metioned tuning constant is presented and illustrated by some simulations using the GLIM4 package. The authors discuss generalisations of this method (i) in the GLM context and (ii) to be joint M-estimation of location and scale. |
Palavras-Chave: |
Simulation. |
Thesagro: |
Estatística; Simulação. |
Thesaurus Nal: |
Least squares; Statistics. |
Categoria do assunto: |
X Pesquisa, Tecnologia e Engenharia |
Marc: |
LEADER 01427naa a2200217 a 4500 001 1006219 005 2010-12-14 008 1994 bl uuuu u00u1 u #d 020 $a3-7908-0793-1 100 1 $aGILCHRIST, R. 245 $aUsing GLIM4 to estimate the tuning constant for Huber's M-estimate of location. 260 $c1994 300 $ap. 251-256. 520 $aProbably the best well known technique for the resistant estimation of a location parameter is Huber's (1964) proposal for an M-estimate of location for the e-contaminated Normal distribution. However, a disadvantage of this technique is the assumption that the amount of contamination, and hence the so called tuning constant is known, a priori. In the present paper the authors present huber's proposal and how it can be implemented using the iteratively te-weighted least dquares (IRLS) procedure used in GLIM4. A proposal for estimating the metioned tuning constant is presented and illustrated by some simulations using the GLIM4 package. The authors discuss generalisations of this method (i) in the GLM context and (ii) to be joint M-estimation of location and scale. 650 $aLeast squares 650 $aStatistics 650 $aEstatística 650 $aSimulação 653 $aSimulation 700 1 $aPORTIDES, G. 773 $tIn: DUTTER, R.; GROSSMANN, W. (Ed.). Compstat: proceedings in computational statistics. [Hildeberg]: Physica-Verlag, 1994.
Download
Esconder MarcMostrar Marc Completo |
Registro original: |
Embrapa Agricultura Digital (CNPTIA) |
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
URL |
Voltar
|
Nenhum registro encontrado para a expressão de busca informada. |
|
Registro completo
Biblioteca(s): |
Catálogo Coletivo de Periódicos Embrapa; Embrapa Agroindústria de Alimentos. |
Identificador: |
3300 |
Data corrente: |
09/05/2002 |
Data da última atualização: |
09/05/2002 |
Código do título: |
5151384 |
Título e Subtítulo: |
FOOD CHEMICAL NEWS GUIDE; CURRENT STATUS OF FOOD ADDITIVES AND COLOR ADDITIVES |
Entidade: |
Food Chemical News |
Local de publicação: |
Washington-DC |
Periodicidade: |
semanal |
Inicio de publicação: |
1964 |
Coleções da unidade: |
Embrapa Agroindústria de Alimentos 1978 20(31, 33-35, 37-41, 47), 1982 24(42), 1983 25(23), 1984 26(36), 1991, 1992, 1993, 1996 Classificação: 664.005 |
|
Fechar
|
|
|