Classe de distribuições de Marshall-Olkin generalizada exponenciada.

Abstract

This work generalizes the family of Marshall-Olkin distributions by adding parameters, making it a new more exible class, creating the new Generalized Exponentialized Marshall-Olkin Weibull distribution (GEMOW). Its probability density function and the associated risk function were studied with promising results. We found some quantities such as moments, moment generating function, quantile function and median, as well Bonferroni and Lorenz curves, for the proposed distribution. We drawed a simulation and we employed the bootstrap resampling procedure for the standard errors of the estimators of the model parameters. We applied the new distribution to magnitudes earthquakes dataset from Fiji archipelago, glass ber resistance dataset to the proposed model, sub-models and competitors distributions. Also it was obtained a regression model for censored data that was applied to data from a study of AIDS, and a Bayesian model implemented for carbon bre data. Comparing with the others distributions, the results demonstrate that GEMOW has superior t to the applied dataset.