Influence of imbalanced datasets in the induction of Full Bayesian Classifiers

This project consists in three main tasks: first, an analysis of the current state of the art in technologies for dealing with class imbalance problems in machine learning algorithms. Second, the analysis of how this problem actually affects a particular class of statistical models, the Bayesian Cla...

Descripción completa

Detalles Bibliográficos
Autor: Morán Jiménez, Daniel
Tipo de recurso: tesis de maestría
Fecha de publicación:2018
País:España
Institución:Universitat Oberta de Catalunya (UOC)
Repositorio:O2, repositorio institucional de la UOC
OAI Identifier:oai:openaccess.uoc.edu:10609/83566
Acceso en línea:http://hdl.handle.net/10609/83566
Access Level:acceso abierto
Palabra clave:Bayesian networks
machine learning
imbalance
aprendizaje automático
desequilibrio
redes bayesianas
aprenentatge automàtic
desequilibri
xarxes bayesianes
Computer algorithms -- TFM
Algorismes computacionals -- TFM
Algoritmos computacionales -- TFM
Descripción
Sumario:This project consists in three main tasks: first, an analysis of the current state of the art in technologies for dealing with class imbalance problems in machine learning algorithms. Second, the analysis of how this problem actually affects a particular class of statistical models, the Bayesian Classifiers, proposing solutions to the particular problems found. And third, to implement a Bayesian Classifier and develop a series of experiments that would support the assertions of the analysis, and shed more light on how this problem can be dealt with.