Estimação Bayesiana para a distribuição Birnbaum- Saunders na presença de dados censurados (Bayesian Estimation for the Birnbaum-Saunders distribution in the presence of censored data)

Fernando Antonio Moala (femoala@fct.unesp.br)1, Jorge Alberto Achcar (achcar@fmrp.usp.br)2, Robson Gimenez (robsongimenezz@hotmail.com)1


1Faculdade de Ciências e Tecnologia - UNESP -Presidente Prudente
2Faculdade de Medicina-USP Ribeirão Preto

This paper appears in: Revista IEEE América Latina

Publication Date: Oct. 2015
Volume: 13,   Issue: 10 
ISSN: 1548-0992


Abstract:
The use of Birnbaum-Saunders distribution can be a good alternative for analyzing data lifetime of equipment. In this work two different prior distributions are used in the estimation of the parameters of the Birnbaum-Saunders distribution under the Bayesian approach and with the presence of type I and II censored data. Assuming a priori dependence between parameters, an alternative prior distribution based on copula functions is proposed. Thus, a study to determine whether the priors lead to the same inference a posteriori is of great practical interest. Two examples are presented to illustrate the proposed methodology and investigated the performance of prior distributions. The Bayesian analysis is performed based on Monte Carlo Markov Chain (MCMC) to generate samples from the posterior distribution.

Index Terms:
Birnbaum-Saunders Distribution, Type I censoring, type II, copula, MCMC   


Documents that cite this document
This function is not implemented yet.


[PDF Full-Text (324)]