• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Doctoral Thesis
DOI
https://doi.org/10.11606/T.45.2021.tde-03052021-165758
Document
Author
Full name
Jairo Arturo Angel Guzman
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2021
Supervisor
Committee
Rocha, Francisco Marcelo Monteiro da (President)
Andrade Filho, Mário de Castro
Paula, Gilberto Alvarenga
Pinheiro, Hildete Prisco
Trinca, Luzia Aparecida
Title in Portuguese
Identificação de modelos lineares mistos gaussianos 
Keywords in Portuguese
Diagnósticos gráficos
Má especificação
Testes
Abstract in Portuguese
O objetivo deste trabalho é avaliar uma possível má especificação de modelos lineares mistos gaussianos. Essa avaliação permite reconhecer quando o modelo é incorretamente especificado e identificar a fonte do erro de especificação, que pode estar na estrutura da média, na estrutura da matriz de covariâncias ou em ambas. Com esse propósito, propomos testes baseados na matriz de informação obtida da função de verossimilhança que, em conjunto com diagnósticos gráficos, são usados na identificação. Um estudo de simulação permite avaliar os testes em quanto ao poder e à taxa do erro tipo I. Concluímos com uma aplicação da estratégia da análise proposta num exemplo com dados reais.
Title in English
Identification of linear Gaussian mixed models
Keywords in English
Graphics diagnostics
Misspecification
Tests
Abstract in English
The objective of this work is to evaluate a possible misspecification in the gaussian linear mixed models. This evaluation makes it possible to recognize when the model is incorrectly specified and to identify the source of the specification error, which can be in the structure of the mean or in the structure of the covariances matrix of the vector of response, or in both cases. For this object, we propose tests based on the Fisher information matrix obtained from the likelihood function in joint together with graphics diagnostics for identification the model. A simulation study allows evaluating the tests in terms of power and type I error rate. We conclude with an application of the proposal developing examples with real data.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Tese_Jairo_USP.pdf (2.80 Mbytes)
Publishing Date
2021-05-04
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.