Learning tractable multidimensional Bayesian network classifiers

Benjumeda Barquita, Marco Alberto and Larrañaga Múgica, Pedro María and Bielza Lozoya, María Concepción (2016). Learning tractable multidimensional Bayesian network classifiers. In: "Eighth International Conference on Probabilistic Graphical Models", 06-09 Sep 2016, Lugano. pp. 13-24.

Description

Title: Learning tractable multidimensional Bayesian network classifiers
Author/s:
  • Benjumeda Barquita, Marco Alberto
  • Larrañaga Múgica, Pedro María
  • Bielza Lozoya, María Concepción
Item Type: Presentation at Congress or Conference (Article)
Event Title: Eighth International Conference on Probabilistic Graphical Models
Event Dates: 06-09 Sep 2016
Event Location: Lugano
Title of Book: Proceedings of the Eighth International Conference on Probabilistic Graphical Models
Date: 2016
Volume: 52
Subjects:
Freetext Keywords: Multidimensional classification; Bayesian network classifiers; MPE complexity; learning from data
Faculty: E.T.S. de Ingenieros Informáticos (UPM)
Department: Inteligencia Artificial
Creative Commons Licenses: Recognition - No derivative works - Non commercial

Full text

[img]
Preview
PDF - Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (214kB) | Preview

Abstract

Multidimensional classification has become one of the most relevant topics in view of the many domains that require a vector of class values to be assigned to a vector of given features. The popularity of multidimensional Bayesian network classifiers has increased in the last few years due to their expressive power and the existence of methods for learning different families of these models. The problem with this approach is that the computational cost of using the learned models is usually high, especially if there are a lot of class variables. Class-bridge decomposability means that the multidimensional classification problem can be divided into multiple subproblems for these models. In this paper, we prove that class-bridge decomposability can also be used to guarantee the tractability of the models. We also propose a strategy for efficiently bounding their inference complexity, providing a simple learning method with an order-based search that obtains tractable multidimensional Bayesian network classifiers. Experimental results show that our approach is competitive with other methods in the state of the art and ensures the tractability of the learned models.

Funding Projects

TypeCodeAcronymLeaderTitle
Government of SpainC080020-09UnspecifiedUnspecifiedCajal Blue Brain Proyect
Madrid Regional GovernmentTIN2013-41592S2013/ICE-2845-CASI-CAM-CMUnspecifiedCajal Blue Brain Proyect
FP7FP7/2007-2013UnspecifiedUnspecifiedHuman Brain Project

More information

Item ID: 46644
DC Identifier: http://oa.upm.es/46644/
OAI Identifier: oai:oa.upm.es:46644
Deposited by: Memoria Investigacion
Deposited on: 18 Oct 2017 11:42
Last Modified: 18 Oct 2017 11:42
  • Logo InvestigaM (UPM)
  • Logo GEOUP4
  • Logo Open Access
  • Open Access
  • Logo Sherpa/Romeo
    Check whether the anglo-saxon journal in which you have published an article allows you to also publish it under open access.
  • Logo Dulcinea
    Check whether the spanish journal in which you have published an article allows you to also publish it under open access.
  • Logo de Recolecta
  • Logo del Observatorio I+D+i UPM
  • Logo de OpenCourseWare UPM