A social-spider optimization approach for support vector machines parameters tuning

Author

Pereira, Danillo R. ; Pazoti, Mario A. ; Pereira, Luis A. M. ; Papa, Joao Paulo

Author_Institution

Inf. Fac. of Presidente Prudente, Univ. of Western Sao Paulo, Presidente Prudente, Brazil

fYear

2014

fDate

9-12 Dec. 2014

Firstpage

1

Lastpage

6

Abstract

The choice of hyper-parameters in Support Vector Machines (SVM)-based learning is a crucial task, since different values may degrade its performance, as well as can increase the computational burden. In this paper, we introduce a recently developed nature-inspired optimization algorithm to find out suitable values for SVM kernel mapping named Social-Spider Optimization (SSO). We compare the results obtained by SSO against with a Grid-Search, Particle Swarm Optimization and Harmonic Search. Statistical evaluation has showed SSO can outperform the compared techniques for some sort of kernels and datasets.

Keywords

learning (artificial intelligence); optimisation; support vector machines; SSO; SVM kernel mapping; SVM-based learning; grid-search comparison; harmonic search comparison; nature-inspired optimization algorithm; particle swarm optimization comparison; social-spider optimization approach; statistical evaluation; support vector machine parameter tuning; Accuracy; Glass; Kernel; Optimization; Sonar; Support vector machines; Training; Evolutionary Computing; Social-Spider Optimization; Support Vector Machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Swarm Intelligence (SIS), 2014 IEEE Symposium on

Conference_Location

Orlando, FL

Type

conf

DOI

10.1109/SIS.2014.7011769

Filename

7011769