预览加载中，请您耐心等待几秒...

改进的遗传算法优化BP神经网络.docx

预览
1/3
2/3
3/3

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

改进的遗传算法优化BP神经网络 Introduction Thebackpropagation(BP)neuralnetworkhasbeenwidelyusedinvariousfieldsduetoitspowerfulnonlinearmappingandlearningability.However,itisachallengingtasktodeterminetheoptimalparametersofBPneuralnetworkforaspecifictask.Traditionaloptimizationmethods,suchasgradientdescentandgeneticalgorithm,havelimitationswhenappliedtotheoptimizationofBPneuralnetwork.Recently,improvedgeneticalgorithmhasbeenproposedtooptimizeBPneuralnetwork,whichhasshownpromisingresultsinmanyapplications. Inthispaper,weintroducetheimprovedgeneticalgorithmforBPneuralnetworkoptimization.WefirstprovideanoverviewofthebackgroundandrelatedworkingeneticalgorithmandBPneuralnetwork.Thenweexplaintheproposedimprovedgeneticalgorithmandcompareitwiththetraditionalmethods.Finally,wedemonstratetheeffectivenessoftheimprovedgeneticalgorithmthroughacasestudy. Background Backpropagationneuralnetworkisatypeoffeedforwardneuralnetwork.Duringthetrainingprocess,thenetworkadjuststheconnectionweightsbetweenneuronstominimizethedifferencebetweentheactualoutputandthedesiredoutput.Thisprocessisknownastheerrorback-propagation(BP)algorithm. Geneticalgorithm(GA)isaheuristicoptimizationalgorithminspiredbythenaturalselectionandevolutionprocess.GAstartswithapopulationofsolutions(individuals),whereeachindividualrepresentsasetofparameters.Thefitnessfunctionevaluatesthequalityofeachindividualinthepopulation,andthegeneticoperator,includingselection,crossover,andmutation,generatesnewindividualsforthenextgeneration. RelatedWork ManyresearchershaveproposedusinggeneticalgorithmtooptimizeBPneuralnetwork.However,traditionalGAhasseveraldrawbacks.Firstly,traditionalGAusuallyrequiresalargenumberofiterationstoconvergetoasatisfactorysolution.Secondly,traditionalGAdoesnotconsiderthecomplexityoftheproblemandmayeasilyfallintoalocaloptimum.Thirdly,traditionalGAhaslowparallelism,whichaffectstheefficiencyofoptimization. Toovercomethesedrawbacks,manyresearchershaveproposedimprovedgeneticalgorithms,suchaselitistGA,adaptiveGA,andhybridGA.Th