Uso do processo gama para dados de sobrevivência

Abstract

In this dissertation, we introduce classical and Bayesian approaches to get inferences on the parameters of interest, considering exponential and Weibull distributions for the lifetimes. For a Bayesian analysis, we assume a gamma process for the individual rates considering type II censoring data and the presence of covariates. We also consider ac- celerated life tests assuming an inverse power law model and an exponential distribution for the lifetimes. The proposed methodology in illustrated in three examples.