支持向量機算法研究及在基因表達數(shù)據(jù)分析中的應用.pdf

1、華中科技大學碩士學位論文支持向量機算法研究及在基因表達數(shù)據(jù)分析中的應用姓名：楊小濤申請學位級別：碩士專業(yè)：計算機應用技術指導教師：劉青20040510華中科技大學碩士學位論文AbstractStatisticallearningtheoryisatheoryofmachinelearninglawdealingwimsmallsamples，andittakesintoaccounttherequirementofthegenerali

2、zationabilityandthemostexcellentanswerinlimitedconditionsBasedonStatisticalLearningTheoryanewmachinelearningmethodsupportvectormachineisputforward，andtherearesomevirtuesindealingwiththeproblemofpatternrecognitionsuchaSth

3、eproblemsofsmallsamples，highdimensionalitynonlinearityMicroarraywillbringaboutarevolutioninbiologyandmedicine，basedonwhichtheexpressionlevelofthousandsofgenesCanbesimultaneouslyobserved，andingenomelevelnatureoflifeisstud

4、iedbyallapproachofsystematicandglobalidea?；datasethassometraits，suchassmallsamples，hi曲dimensionalitynonlinearitytooItisachallengeforsometraditionalmachinelearningmethodsandtheanalyticalmethodisahottopicinbioinformaticsn

5、owTheoryandmethodofsupportvectormachinearestudiedintypicalgeneexpressiondatasetsIntheory，severalkeyissuesinstatisticaAlearningtheoryandsupportvectormachinealediscussed；atrainingalgorithmofsupportvectormachinesequentialmi

6、nimaloptimizationisgiven，andaimprovedversionofitisputforwardinunequalsamplenumberbetweenclassesthemethodisprovedtobefeasiblethroughsimulativedata；asoftwareGeneSVMbasedonsupportvectormachineforgeneexpressiondataisachieved

7、andaclassifiermodelisbroughtforward：firstgenesubsetisextractedbythemethodofsignaltonoiseratio；thendatasetisnormalizedbyrainmaxmethod；intheendaclassifierisbuiltbythemethodofSVM，aclassifiermodelbasedonabovetheoriesissucces