Publikationsansicht

RBF CENTERS INITIALIZATION USING FUZZY CLUSTERING TECHNIQUE FOR FUNCTION APPROXIMATION PROBLEMS

Abstract
In this paper, a new algorithm for RBF centers initialization for functional approximation problems is proposed. The design of the algorithm is inspired in previous clustering algorithms but adding some features to adapt the algorithm to the characteristics of our problem. We compare the results provided by our new algorithm with the results provided by the algorithms it is based on.. Neural networks, radial basis functions, function approximation, clustering

Details der Publikation
Archiv RePEc (Germany)
Typ article