Métodos de análisis de conglomerados difusos con entropía relativa de Rényi
Loading...
Download
Official URL
Full text at PDC
Publication date
2025
Defense date
19/04/2024
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Complutense de Madrid
Citation
Abstract
El análisis de conglomerados, como parte del análisis multivariante de datos, es un conjunto de métodos los cuales tienen como idea principal crear grupos o clústeres con dos condiciones: la primera es que todos los objetos de un grupo han de ser similares u homogéneos, y la segunda es que los objetos han de ser diferentes o heterogéneos cuando pertenezcan a diferentes grupos. Esta tesis se va a centrar en los métodos basados en centroides (medias ponderadas), los cuales asumen el conocimiento a priori del número óptimo de clases, y que recolocan dichos centroides durante un proceso iterativo hasta encontrar una solución o partición del conjunto de datos que sea estable o fija. Como objetivo principal de esta tesis se persigue el desarrollo de una nueva función objetivo y unos nuevos métodos análisis de conglomerados que obtenga mejores resultados que los existentes en la literatura...
Cluster analysis, as part of multivariate data analysis, is a set of methods whose main idea is to create groups or clusters with two conditions: the first is that all objects in a group must be similar or homogeneous, and the second is that objects must be different or heterogeneous when they belong to different groups. This thesis will focus on methods based on centroids (weighted averages), which assume a priori knowledge of the optimal number of classes, and which relocate these centroids during an iterative process until a stable or fixed solution or partition of the dataset is found.The main objective of this thesis is the development of a new objective function and a new cluster analysis method that obtains better results than those existing in the literature...
Cluster analysis, as part of multivariate data analysis, is a set of methods whose main idea is to create groups or clusters with two conditions: the first is that all objects in a group must be similar or homogeneous, and the second is that objects must be different or heterogeneous when they belong to different groups. This thesis will focus on methods based on centroids (weighted averages), which assume a priori knowledge of the optimal number of classes, and which relocate these centroids during an iterative process until a stable or fixed solution or partition of the dataset is found.The main objective of this thesis is the development of a new objective function and a new cluster analysis method that obtains better results than those existing in the literature...
Description
Tesis inédita de la Universidad Complutense de Madrid, Facultad de Ciencias Matemáticas, leída el 19/04/2024