Please use this identifier to cite or link to this item:

Title: Training Support Vector Machines with an Heterogeneous Particle Swarm Optimizer
Authors: Silva, Arlindo
Ana, neves
Gonçalves, Teresa
Keywords: support vector machines,
non PSD kernels
particle swarm optimization
heterogeneous particle swarms
Issue Date: 2013
Publisher: Springer Berlin Heidelberg
Citation: A. Silva, A. Neves and T. Gonçalves. Training support vector machines with an heterogeneous particle swarm optimizer. In ICANNGA’13 – Adaptive and Natural Computing Algo- rithms, volume 7824 of Lecture Notes in Computer Science, pages 100–109. Springer Berlin Heidelberg, April 2013
Abstract: Support vector machines are classification algorithms that have been successfully applied to problems in many different areas. Re- cently, evolutionary algorithms have been used to train support vector machines, which proved particularly useful in some multi-objective for- mulations and when indefinite kernels are used. In this paper, we propose a new heterogeneous particle swarm optimization algorithm, called scout- ing predator-prey optimizer, specially adapted for the training of support vector machines. We compare our algorithm with two other evolutionary approaches, using both positive definite and indefinite kernels, on a large set of benchmark problems. The experimental results confirm that the evolutionary algorithms can be competitive with the classic methods and even superior when using indefinite kernels. The scouting predator-prey optimizer can train support vector machines with similar or better classi- fication accuracy than the other evolutionary algorithms, while requiring significantly less computational resources.
Type: article
Appears in Collections:INF - Artigos em Livros de Actas/Proceedings

Files in This Item:

File Description SizeFormat
silva13b-training.pdf319.3 kBAdobe PDFView/OpenRestrict Access. You can Request a copy!
FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpaceOrkut
Formato BibTex mendeley Endnote Logotipo do DeGóis 

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.


Dspace Dspace
DSpace Software, version 1.6.2 Copyright © 2002-2008 MIT and Hewlett-Packard - Feedback
UEvora B-On Curriculum DeGois