• 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
 
 
Master's Dissertation
DOI
10.11606/D.55.2018.tde-19032018-164343
Document
Author
Full name
Daniela Brassolatti
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 1997
Supervisor
Committee
Achcar, Jorge Alberto (President)
Rodrigues, Josemar
Wechsler, Sergio
Title in Portuguese
Uso de Métodos Bayesianos na Análise de Dados de Confiabilidade de Software Considerando Tempos entre Falhas
Keywords in Portuguese
Não disponível
Abstract in Portuguese
Nesta dissertação de mestrado, apresentamos análises Clássica e Bayesiana para os principais modelos de Estratégia tipo I, estratégia de modelos de confiabilidade de software que modelam os tempos entre falhas do software. Na análise Clássica, estimadores pontuais e intervalos de confiança são encontrados usando métodos assintóticos. Na análise Bayesiana, considerando densidades a priori informativas para os parâmetros dos modelos, determinamos os resumos a posteriori, utilizando os métodos de simulação Gibbs Sampling e Metrópolis Hastings. Em particular, consideramos diferentes densidades a priori para os parâmetros do modelo de Jelinski e Moranda (1972) (um dos primeiros modelos de confiabilidade de software desenvolvido) e verificamos a consequência de uma reparametrização para esse modelo. Também, apresentamos a técnica das distribuições preditivas condicionais ordenadas (CPO) para selecionar o melhor modelo dentre os modelos analisados. Finalizamos, ilustrando os métodos propostos através de um exemplo prático
Title in English
Not available
Keywords in English
Not available
Abstract in English
In this dissertation we preseut Classical and Bayesian analyses for the most important models of strategy of type-I used in software reliability to model times between failures. In the Classical approach, point estimators and confidence intervals are obtained using assymptotical methods. In the Bayesian approach, considering informative prior densities for the parameters of the models, we obtain posterior surmnaries of interest, using the simulatim algorithms Gibbs Sampling and Metrópolis Hastings. In special, we consider different prior densities for the parameters of the Jelinki and Moranda (1972) model and we also check the effects of a reparametrization in the obtained inferences. We also consider the technique of ordinated conditional predictive distributions (CPO) to select the best model among all considered models. We conclude the work, presenting a practical example as a numerical illustration of the proposed methodology.
 
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.
DanielaBrassolatti.pdf (91.32 Mbytes)
Publishing Date
2018-03-19
 
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 © 2001-2021. All rights reserved.