Nuevos modelos de distribuciones para datos de conteo

  1. CUEVA LÓPEZ, VALENTINA
Supervised by:
  1. José Rodríguez Avi Director
  2. María José Olmo Jiménez Co-director

Defence university: Universidad de Jaén

Fecha de defensa: 02 May 2023

Committee:
  1. Antonio Pascual Acosta Chair
  2. Emilio Lozano Aguilera Secretary
  3. Rafaela Osuna Gómez Committee member

Type: Thesis

Teseo: 819998 DIALNET

Abstract

In this paper, two new discrete distributions for count data, belonging to the Pearson family of discrete distributions, are described and developed. In both cases, an exhaustive study of the resulting distribution is made, the parameters are analyzed and interpreted, the corresponding estimation parameters are developed, and they are compared with existing distributions in the literature. It is noteworthy that they can be applied in multiple fields and solve some problems present in other models, such as the case of the indeterminacy of the variance in the Waring generalized univariate distribution, the modeling of overdispersed data with a mode other than 1, or the possibility of Obtain over and underdispersed models, with the advantage of having clean expressions of the parameters. Finally, the models developed are applied to data from the economy, agriculture, literature or education.