摘要
Basedonimmuneclusteringandevolutionaryprogramming(EP),ahybridalgorithmtotraintheRBFnetworkisproposed.AnimmunefuzzyC-meansclusteringalgorithm(IFCM)isusedtoadaptivelyspecifytheamountandinitialpositionsoftheRBFcentersaccordingtoinputdataset;thentheRBFnetworkistrainedwithEPthattendstoglobaloptima.TheapplicationofthehybridalgorithminmultluserdetectionproblemdemonstratesthattheRBFnetworktrainedwiththealgorithmhassimplenetworkstructurewithgoodgeneralizationability.
出版日期
2004年03月13日(中国期刊网平台首次上网日期,不代表论文的发表时间)