Please use this identifier to cite or link to this item: http://repositorio.udec.cl/jspui/handle/11594/3535
Title: Trapezoidal Kumaraswamy Distribution Distribución Kumaraswamy Trapezoidal
Authors: Figueroa Zúñiga, Jorge Isaac , supervisora de grado
Sanhueza Parkes, Rodrigo Alonso
Keywords: Algorítmos de Expectación-Maximización
Distribución Kuamaraswamy
Issue Date: 2018
Publisher: Universidad de Concepción. Facultad de Ciencias Físicas y Matemáticas .Departamento de Estadística
Abstract: n the year 1980 Poondi Kumaraswamy proposed the Kuamaraswamy distribution, which is very similar to the beta distribution (it is also restricted to the interval (0,1)), but it has a great advantage over it, which is to have a distribution function accumulated in a closed form which is more beneficial for intensive calculation activities such as simulation modeling and estimation of models by methods based on simulation. The problem of this distribution and its extensions proposed in the following years is that they have not been able to adjust the data that sometimes are concentrated in each of the extremes or both ends independently. This work has the purpose of showing the proposal of a new distribution, which has been called trapezoidal kumaraswamy distribution which has been originated by mixing the Kumaraswamy distribution and the Beta distribution, making the tails of the density function more flexible in one of the extremes or in both of them independently with which a greater adjustment of the data is achieved. We can appreciate the properties of this new model and the estimation of parameters. Finally, a simulation study and an application of real data is presented, with the intention of showing the best adjustment obtained.
Description: Magíster en Estadística Universidad de Concepción 2018
URI: http://repositorio.udec.cl/jspui/handle/11594/3535
Appears in Collections:Estadística - Tesis Magister

Files in This Item:
File Description SizeFormat 
Tesis_Trapezoidal_Kumaraswamy_.Image.Marked.pdf323,51 kBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.