Bio-Inspired Systems: Computational and Ambient Intelligence
Lecture Notes in Computer Science, 2009, Volume 5517/2009, 383-390, DOI: 10.1007/978-3-642-02478-8_48

Parallelizing the Design of Radial Basis Function Neural Networks by Means of Evolutionary Meta-algorithms

M. G. Arenas, E. Parras-Gutiérrez, V. M. Rivas, P. A. Castillo, M. J. Del Jesus and J. J. Merelo

View Related Documents

Abstract

This work introduces SymbPar, a parallel meta-evolutionary algorithm designed to build Radial Basis Function Networks minimizing the number of parameters needed to be set by hand. Parallelization is implemented using independent agents to evaluate every individual. Experiments over classifications problems show that the new method drastically reduces the time took by sequential algorithms, while maintaining the generalization capabilities and sizes of the nets it builds.

Keywords  Neural networks - evolutionary algorithms - parallelization - meta-evolution

Fulltext Preview

Image of the first page of the fulltext document