Please use this identifier to cite or link to this item: http://repositorio.ufla.br/jspui/handle/1/11100
Title: Novo estimador de cumeeira de Rao com aplicação em seleção genômica
Other Titles: New estimator of ridge Rao with application in genomic selection
Authors: Chaves, Lucas Monteiro
Ferreira, Daniel Furtado
Jaques, Devanil
Silva, Fabyano Fonseca e
Oliveira, Marcos Santos de
Keywords: Análise Multivariada
Estimadores viesados
Multicolinearidade
Ridge Regression
Multivariate Analysis
Biased estimators
Multicollinearity
Issue Date: 28-Apr-2016
Publisher: Universidade Federal de Lavras
Citation: COSTA, L. A. da. Novo estimador de cumeeira de Rao com aplicação em seleção genômica. 2015. 126 p. Tese (Doutorado em Estatística e Experimentação Agropecuária)-Universidade Federal de Lavras, Lavras, 2015.
Abstract: A new ridge estimator is obtained from a geometric approach of the general theory of the ridge estimators. Obtained from the intersection of two orthogonal ellipsoids and having a simple algebraic expression this estimator is quite natural. Examples are used to suggest that curves ridge from this estimator stabilize faster than the usual ridge estimators. This fact may facilitate the estimation of great value for the parameter, a central area of research in theory. Expressions were obtained for the total variance and the mean quadratic error. The performance of the proposed estimator was studied by applying a set of data already analyzed in the literature by using the usual ridge estimator. Data analysis of genomic selection on the pH response of pork at 45 minutes and 24 hours after slaughter. In all applications the results obtained by the proposed estimator are compared with the results obtained by the usual ridge and Mayer and Wilke (1973) estimator.
URI: http://repositorio.ufla.br/jspui/handle/1/11100
Appears in Collections:Estatística e Experimentação Agropecuária - Doutorado (Teses)

Files in This Item:
File Description SizeFormat 
TESE_Novo estimador de cumeeira de Rao com aplicação em seleção.pdf1,69 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.