Zero-Truncated Poisson Exponentiated Gamma Distribution: Application and Estimation Methods

• Autor(a) principal: Aguilar, Guilherme A. S.
• Data de Publicação: 2019
• Outros Autores: Moala, Fernando A. [UNESP], Cordeiro, Gauss M.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• DOI: 10.1007/s42519-019-0059-2
• Texto Completo: http://dx.doi.org/10.1007/s42519-019-0059-2; http://hdl.handle.net/11449/186141
• topic: Exponentiated gamma; Maximum likelihood; Zero-truncated Poisson; Bayesian estimation

We define a new two-parameter lifetime model called the zero-truncated Poisson exponentiated gamma distribution. The hazard function of the new distribution has monotonic and non-monotonic shapes, which allows to fit more dispersed data. The proposed distribution has the advantage of having only two parameters and consequently provides an easier way for estimating the model parameters differently from other distributions with three or more parameters. We obtain some of its structural properties. Several methods of estimation of the model parameters are presented. A simulation with ten different models of estimators was performed in order to compare the performance and accuracy of such proposed estimators. An application to real data was included.