Bibliographic citations
This is an automatically generated citacion. Modify it if you see fit
Benites, L., (2018). Finite mixtures of regression models [Tesis, Universidade Estadual de Campinas]. http://renati.sunedu.gob.pe/handle/sunedu/952729
Benites, L., Finite mixtures of regression models [Tesis]. BR: Universidade Estadual de Campinas; 2018. http://renati.sunedu.gob.pe/handle/sunedu/952729
@phdthesis{renati/1362,
title = "Finite mixtures of regression models",
author = "Benites Sánchez, Luis Enrique",
publisher = "Universidade Estadual de Campinas",
year = "2018"
}
Title: Finite mixtures of regression models
Other Titles: Mixtura finita de los modelos de regresión
Authors(s): Benites Sánchez, Luis Enrique
Advisor(s): Bolfarine, Heleno; Lachos Dávila, Víctor Hugo
Keywords: Estadística matemática; Modelos matemáticos; Modelos lineales (Estadística); Análisis de regresión; Análisis multivariante
OCDE field: http://purl.org/pe-repo/ocde/ford#1.01.03
Issue Date: May-2018
Institution: Universidade Estadual de Campinas
Abstract: Esta disertación consta de tres artículos que proponen extensiones de mezclas finitas en modelos de regresión. Aquí consideramos una clase flexible de distribuciones univariadas y multivariadas, que permiten el modelado adecuado de datos asimétricos que tienen multimodalidad, colas pesadas y observaciones periféricas. Esta clase tiene casos especiales tales como distribuciones contaminadas skew-normal, skew-t, skew-slash y skew normal, así como casos simétricos. Inicialmente, se propone un modelo basado en la suposición de que los errores siguen una mezcla finita de mezcla de escala de distribución asimétrica normal (FM-SMSN) en lugar de la distribución normal convencional. A continuación, tenemos un modelo de regresión censurado donde consideramos que el error sigue una mezcla finita de mezcla de escala de distribución normal (SMN). A continuación, proponemos un modelo de regresión censurado donde consideramos que el error sigue una mezcla finita de mezcla de escala de distribución normal (SMN). Finalmente, consideramos una mezcla finita de regresión multivariada donde el error tiene una distribución multivariada de SMSN. Para todos los modelos propuestos, se desarrollaron dos paquetes R, que se informan en el apéndice.
This dissertation consists of three articles, proposing extensions of finite mixtures in regression models. Here we consider a flexible class of both univariate and multivariate distributions, which allow adequate modeling of asymmetric data that have multimodality, heavy tails and outlying observations. This class has special cases such as skew-normal, skew-t, skew-slash and skew normal contaminated distributions, as well as symmetric cases. Initially, a model is proposed based on the assumption that the errors follow a finite mixture of scale mixture of skew-normal (FM-SMSN) distribution rather than the conventional normal distribution. Next, we have a censored regression model where we consider that the error follows a finite mixture of scale mixture of normal (SMN) distribution. Next, we propose a censored regression model where we consider that the error follows a finite mixture of scale mixture of normal (SMN) distribution. Finally, we consider a finite mixture of multivariate regression where the error has a multivariate SMSN distribution. For all proposed models, two R packages were developed, which are reported in the appendix.
This dissertation consists of three articles, proposing extensions of finite mixtures in regression models. Here we consider a flexible class of both univariate and multivariate distributions, which allow adequate modeling of asymmetric data that have multimodality, heavy tails and outlying observations. This class has special cases such as skew-normal, skew-t, skew-slash and skew normal contaminated distributions, as well as symmetric cases. Initially, a model is proposed based on the assumption that the errors follow a finite mixture of scale mixture of skew-normal (FM-SMSN) distribution rather than the conventional normal distribution. Next, we have a censored regression model where we consider that the error follows a finite mixture of scale mixture of normal (SMN) distribution. Next, we propose a censored regression model where we consider that the error follows a finite mixture of scale mixture of normal (SMN) distribution. Finally, we consider a finite mixture of multivariate regression where the error has a multivariate SMSN distribution. For all proposed models, two R packages were developed, which are reported in the appendix.
Link to repository: http://renati.sunedu.gob.pe/handle/sunedu/952729
Discipline: Estadística
Grade or title grantor: Universidade de São Paulo. Instituto de Matemática e Estatística
Grade or title: Doctor en Ciencias
Register date: 27-Apr-2020
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Benites_Sanchez_LE.d.pdf Restricted Access | Tesis (abierta en repositorio de origen) | 2.01 MB | Adobe PDF | View/Open Request a copy |
Autorizacion.pdf Restricted Access | Formato de autorización | 1.12 MB | Adobe PDF | View/Open Request a copy |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.