Técnicas de formación de grupos: Métodos de particionamiento
Loading...
Identifiers
Publication date
Authors
Advisors
Tutors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
[ES] El análisis cluster es un conjunto de técnicas de análisis multivariante que permiten
clasificar conjuntos de datos en grupos, de forma que los individuos dentro de cada grupo
presenten cierto grado de homogeneidad respecto de las variables observables. En este
trabajo hemos revisado de forma teórica los métodos de K−medias y de mixturas finitas
para la agrupación de un conjunto de datos, y hemos realizado un análisis comparativo
de ambos métodos. En el primer capítulo se ofrece una introducción al análisis cluster y,
en particular, a los métodos de particionamiento. En el segundo y tercer capítulos se hizo
una revisión teórica de los algoritmo de K−medias y mixturas finitas, respectivamente,
y se ilustran los algoritmos mediante un ejemplo con datos reales. En el cuarto, y último
capítulo, hemos realizado varias simulaciones que nos permitieron comparar los algoritmos
de K−medias y mixturas finitas y hemos podido ver en qué circunstancias es más adecuado
aplicar cada uno de los métodos.
[EN] Cluster analysis is a set of multivariate analysis techniques that allow data sets to be classified into groups, so that the individuals within each group present a certain degree of homogeneity with respect to the observable variables. In this work we have theoretically reviewed the K−means and finite mixture methods for grouping a data set, and we have performed a comparative analysis of both methods. The first chapter provides an introduction to cluster analysis and, in particular, to partitioning methods. In the second and third chapters it was made a theoretical review of the algorithms of K−means and finite mixtures, and the algorithms are illustrated by an example with real data. In the fourth, and last chapter, we have carried out several simulations that allowed us to compare the algorithms of K−means and finite mixtures and we have been able to see in which circumstances it is more appropriate to apply each of the methods.
[EN] Cluster analysis is a set of multivariate analysis techniques that allow data sets to be classified into groups, so that the individuals within each group present a certain degree of homogeneity with respect to the observable variables. In this work we have theoretically reviewed the K−means and finite mixture methods for grouping a data set, and we have performed a comparative analysis of both methods. The first chapter provides an introduction to cluster analysis and, in particular, to partitioning methods. In the second and third chapters it was made a theoretical review of the algorithms of K−means and finite mixtures, and the algorithms are illustrated by an example with real data. In the fourth, and last chapter, we have carried out several simulations that allowed us to compare the algorithms of K−means and finite mixtures and we have been able to see in which circumstances it is more appropriate to apply each of the methods.
Description
Traballo Fin de Grao en Matemáticas. Curso 2019-2020
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Atribución-NoComercial-CompartirIgual 4.0 Internacional



