Feature selection for Support Vector Machines via Mixed Integer Linear Programming
Author
dc.contributor.author
Maldonado, Sebastián
Author
dc.contributor.author
Pérez, Juan
es_CL
Author
dc.contributor.author
Weber, Richard
es_CL
Author
dc.contributor.author
Labbé, Martine
es_CL
Admission date
dc.date.accessioned
2014-12-30T13:29:12Z
Available date
dc.date.available
2014-12-30T13:29:12Z
Publication date
dc.date.issued
2014
Cita de ítem
dc.identifier.citation
Information Sciences 279 (2014) 163–175
en_US
Identifier
dc.identifier.other
DOI: 10.1016/j.ins.2014.03.110
Identifier
dc.identifier.uri
https://repositorio.uchile.cl/handle/2250/126859
General note
dc.description
Artículo de publicación ISI
en_US
Abstract
dc.description.abstract
The performance of classification methods, such as Support Vector Machines, depends
heavily on the proper choice of the feature set used to construct the classifier. Feature
selection is an NP-hard problem that has been studied extensively in the literature. Most
strategies propose the elimination of features independently of classifier construction by
exploiting statistical properties of each of the variables, or via greedy search. All such strategies
are heuristic by nature. In this work we propose two different Mixed Integer Linear
Programming formulations based on extensions of Support Vector Machines to overcome
these shortcomings. The proposed approaches perform variable selection simultaneously
with classifier construction using optimization models. We ran experiments on real-world
benchmark datasets, comparing our approaches with well-known feature selection techniques
and obtained better predictions with consistently fewer relevant features.
en_US
Patrocinador
dc.description.sponsorship
Support from the Institute of Complex Engineering Systems (ICM: P-05-004-F, CONICYT: FBO16)