Strategies for piecing-together local-to-global markov network learning algorithms

We introduce in this work a set of strategies for improving the piecing-together step in local-to-global Markov networks structure learning algorithms. For Markov networks, Local-to-global algorithms decompose the problem of learning a complete independence structure with n variables into n independ...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Schlüter, Federico, Bromberg, Facundo, Abraham, Laura
Formato:	Objeto de conferencia
Lenguaje:	Inglés
Publicado:	2011
Materias:	Ciencias Informáticas Markov networks structure learning independence-based global learning
Acceso en línea:	http://sedici.unlp.edu.ar/handle/10915/125248
Aporte de:	SEDICI (UNLP) de Universidad Nacional de La Plata

id	I19-R120-10915-125248
record_format	dspace
spelling	I19-R120-10915-1252482023-05-08T17:23:03Z http://sedici.unlp.edu.ar/handle/10915/125248 issn:1850-2784 Strategies for piecing-together local-to-global markov network learning algorithms Schlüter, Federico Bromberg, Facundo Abraham, Laura 2011-08 2011 2021-09-21T14:07:17Z en Ciencias Informáticas Markov networks structure learning independence-based global learning We introduce in this work a set of strategies for improving the piecing-together step in local-to-global Markov networks structure learning algorithms. For Markov networks, Local-to-global algorithms decompose the problem of learning a complete independence structure with n variables into n independent Markov blanket learning problems. On a second step these algorithms piece-together all the learned Markov blankets into a global structure using an \OR rule". Insucient data may result in incorrect learning of Markov blankets, with con in their decision on edge inclusion when, for two variables X and Y , X is in the blanket of Y , but Y is not in the blanket of X. In such cases the \OR rule" always decides to add the edge, making mistakes when such edge does not exist. Our contribution is alternative strategies. The first alternative is based on the \AND rule" which proposes to add an edge between two variables X and Y to the global structure if they mutually belong to its respective Markov blankets. The other alternative rule is based on the probability of the edges and aims to solve an inconsistency by comparing the probability of edge existence with the probability of edge absence, and taking the more probable for deciding to add or remove such edge. At the end of the work, we show that inconsistencies are an important source of errors in these algorithms, and demonstrate empirically interesting improvements in the quality of learned structures, using this new piecing-together alternative instead of the basic \OR rule". Sociedad Argentina de Informática e Investigación Operativa Objeto de conferencia Objeto de conferencia http://creativecommons.org/licenses/by-nc-sa/4.0/ Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0) application/pdf 96-107
institution	Universidad Nacional de La Plata
institution_str	I-19
repository_str	R-120
collection	SEDICI (UNLP)
language	Inglés
topic	Ciencias Informáticas Markov networks structure learning independence-based global learning
spellingShingle	Ciencias Informáticas Markov networks structure learning independence-based global learning Schlüter, Federico Bromberg, Facundo Abraham, Laura Strategies for piecing-together local-to-global markov network learning algorithms
topic_facet	Ciencias Informáticas Markov networks structure learning independence-based global learning
description	We introduce in this work a set of strategies for improving the piecing-together step in local-to-global Markov networks structure learning algorithms. For Markov networks, Local-to-global algorithms decompose the problem of learning a complete independence structure with n variables into n independent Markov blanket learning problems. On a second step these algorithms piece-together all the learned Markov blankets into a global structure using an \OR rule". Insucient data may result in incorrect learning of Markov blankets, with con in their decision on edge inclusion when, for two variables X and Y , X is in the blanket of Y , but Y is not in the blanket of X. In such cases the \OR rule" always decides to add the edge, making mistakes when such edge does not exist. Our contribution is alternative strategies. The first alternative is based on the \AND rule" which proposes to add an edge between two variables X and Y to the global structure if they mutually belong to its respective Markov blankets. The other alternative rule is based on the probability of the edges and aims to solve an inconsistency by comparing the probability of edge existence with the probability of edge absence, and taking the more probable for deciding to add or remove such edge. At the end of the work, we show that inconsistencies are an important source of errors in these algorithms, and demonstrate empirically interesting improvements in the quality of learned structures, using this new piecing-together alternative instead of the basic \OR rule".
format	Objeto de conferencia Objeto de conferencia
author	Schlüter, Federico Bromberg, Facundo Abraham, Laura
author_facet	Schlüter, Federico Bromberg, Facundo Abraham, Laura
author_sort	Schlüter, Federico
title	Strategies for piecing-together local-to-global markov network learning algorithms
title_short	Strategies for piecing-together local-to-global markov network learning algorithms
title_full	Strategies for piecing-together local-to-global markov network learning algorithms
title_fullStr	Strategies for piecing-together local-to-global markov network learning algorithms
title_full_unstemmed	Strategies for piecing-together local-to-global markov network learning algorithms
title_sort	strategies for piecing-together local-to-global markov network learning algorithms
publishDate	2011
url	http://sedici.unlp.edu.ar/handle/10915/125248
work_keys_str_mv	AT schluterfederico strategiesforpiecingtogetherlocaltoglobalmarkovnetworklearningalgorithms AT brombergfacundo strategiesforpiecingtogetherlocaltoglobalmarkovnetworklearningalgorithms AT abrahamlaura strategiesforpiecingtogetherlocaltoglobalmarkovnetworklearningalgorithms
_version_	1765659840543195136

Strategies for piecing-together local-to-global markov network learning algorithms

Ejemplares similares