
handle: 10835/19116
En este trabajo se estudiarán desde un punto de vista teórico las redes bayesianas, sus propiedades y algunos algoritmos de aprendizaje e inferencia. Además, se llevará a la práctica todo esto a través del estudio y aplicación sobre conjuntos de datos de dis- tinto tipo mediante ciertos paquetes disponibles en R para la modelización con redes bayesianas, que serán nuestro objeto de estudio. Los paquetes que estudiaremos serán: bnlearn, catnet, deal, gRain y MoTBFs. De este modo, en el segundo capítulo introduciremos el concepto de red bayesiana como forma de modelar probabilísticamente la incertidumbre. Daremos la definición y propiedades de determinados conceptos necesarios para llevar a cabo la modelización de datos a través de redes bayesianas. Tras esto, en el tercer capítulo, daremos comienzo a una exposición de las etapas involucradas en la modelización con redes bayesianas, donde hablaremos de las técnicas y algoritmos usados en los procesos de aprendizaje, estructural y paramétrico, e inferencia. El cuarto capítulo se centrará en aplicar los algoritmos estudiados con varios conjuntos de datos de distinta naturaleza, y exponer las diversas funciones disponibles en cada uno de los paquetes de R analizados. El objetivo de esta parte es que sirva al lector como guía de consulta a la hora de realizar tareas de aprendizaje e inferencia mediante redes bayesianas. Se finaliza con una discusión y comparativa sobre los paquetes analizados y las principales conclusiones del trabajo. In this work, we will study Bayesian networks from a theoretical perspective, exploring their properties and various learning and inference algorithms. We will also apply this knowledge in practice by working with datasets of different types and spe- cific R packages for Bayesian network modeling, which will be the focus of our study. The packages we will investigate include bnlearn, catnet, deal, gRain, and MoTBFs. In the second chapter, we will introduce the concept of Bayesian networks as a way to probabilistically model uncertainty. We will provide definitions and discuss the properties of key concepts necessary for data modeling with Bayesian networks. Following that, in the third chapter, we will delve into the stages involved in Bayesian network modeling, discussing the techniques and algorithms used in the structural and parametric learning processes, as well as inference. The fourth chapter will focus on applying the studied algorithms to various datasets of different nature, showcasing the functions available in each of the analyzed R packages. The aim of this section is to serve as a reference guide for readers when performing learning and inference tasks using Bayesian networks. We will conclude with a discussion and comparison of the analyzed packages and the main findings of the study.
gRain, Redes bayesianas, Deal, Catnet, MoTBFs, Bnlearn
gRain, Redes bayesianas, Deal, Catnet, MoTBFs, Bnlearn
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 0 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
