支持向量机及其在铁路工程中的应用研究(I)

《支持向量机及其在铁路工程中的应用研究(I)》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、西南交通大学硕士学位论文支持向量机及其在铁路工程中的应用研究姓名：蒋学申请学位级别：硕士专业：道路与铁道工程指导教师：李远富20081001西南交通大学硕士研究生学位论文第1I页AbstractSupportVectorMachine(SVM)basedontheSLT(statisticallearningtheory)isanewmachinelearningmethod，whichwasdevelopedbyVapnikandhisteamin1995；itembodiesthetheoryofstructur

2、eriskminimization(SRM)andCansolvetheproblemcharacterizedbynonlinear,highdimension，smallsampleandlocalmi‘ni‘mi。zi‘ngperfectly．SVMhasbecomethehotspotinthefieldofmachinelearningbecauseofitsexcellentlearningperformance，SOitwasappliedsuccessfullyinmanyengineeringfie

3、lds．ThisthesisappliedSVMtotherailwayengineefing，becauseithasbetterlearningfeatureandfutureapplyvaluewhichishopedtosolvemuchprobleminthedatamining(DM)．ThefollowingworkWasconductedinthisthesis：．1．Expatiatedandcomparedthecommonmethodsinthepassengerpredictionandinv

4、estmentprediction．Detailedlydeducedthetraininganddecision·-makingprocessofSVMfromlinearSVMtoNon··linearSVMandsumthetrainingalgorithm．2．Summarizedthetheoryofneuralnetwork，constructedthepassengervolumepredictionmodelbasedonBPandrailwayinvestmentpredictionmodelbas

5、edonRBFbyprogrammingonMATLABlanguage．3．ConstructedtheregressionmodelonthebaseoftheoryofSVM，andthenappliedthismodelonthepredictionofcityrailwaypassengervolumebyLIBSVM．ThencomparedtheresultsbetweenSVMandBP,resultindicatethatSVMismoreprecisethanNNinthesituationofs

6、mallsample．4．Predictionoftheinvestmentisthehotspotofrailwayproject，constructedthesupportvectorregressionmodelbyLIBSVM，andapplieditontheTBMrailwaytunnelcostprediction，resultshowthatSVMhaveabetterperformancethanNNinthesituationofhighdimension．111eresultsshowthatS

7、VMmethodiSbetterthanneuralnetwork．ItiSbelievedthatUsingSVMtheorytOsolveregressionproblemisamethodwithpromisingprospect．Atlast，apersonalpreviewoffurthertasksintheresearchrealmsofneuralnetworkandsupportvectormachineiSpresented．西南交通大学硕士研究生学位论文第1II页-_●_●■●■_●I●____

8、●■●■■■■●■●■_●■●■■●■●mn-ml●■l●●●■●_■●■●■l_■●■_●■_■_●_■■●_●●l●■●■____●Keywords：supportvectormachine；neuralnetwork；regressionmodel；railwaypassengervolume；investmentprediction西南