Optimized FCM-based radial basis function neural networks: A comparative analysis of LSE and WLSE method

Wook Dong Kim*, Sung Kwun Oh, Wei Huang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

In this paper, we introduce a new architecture of optimized FCM-based Radial Basis function Neural Network by using space search algorithm and discuss its comprehensive design methodology. As the consequent part of rules of the FCM-based RBFNN model, four types of polynomials are considered. The performance of the FCM-based RBFNN model is affected by some parameters such as the number of cluster and the fuzzification coefficient of the fuzzy clustering (FCM) and the order of polynomial standing in the consequent part of rules, we are required to carry out parametric optimization of network. The space evolutionary algorithm(SEA) being applied to each receptive fields of FCM-based RBFNN leads to the selection of preferred receptive fields with specific local characteristics available within the FCM-based RBFNN. The performance of the proposed model and the comparative analysis between WLSE and LSE are illustrated with by using two kinds of representative numerical dataset.

Original languageEnglish
Title of host publicationAdvances in Neural Networks - ISNN 2010 - 7th International Symposium on Neural Networks, ISNN 2010, Proceedings
Pages207-214
Number of pages8
EditionPART 1
DOIs
Publication statusPublished - 2010
Externally publishedYes
Event7th International Symposium on Neural Networks, ISNN 2010 - Shanghai, China
Duration: 6 Jun 20109 Jun 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume6063 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Symposium on Neural Networks, ISNN 2010
Country/TerritoryChina
CityShanghai
Period6/06/109/06/10

Keywords

  • Fuzzy C-means clustering
  • Machine learning data
  • Radial basis function neural network
  • Space evolutionary algorithm

Fingerprint

Dive into the research topics of 'Optimized FCM-based radial basis function neural networks: A comparative analysis of LSE and WLSE method'. Together they form a unique fingerprint.

Cite this