生物信息數(shù)據(jù)挖掘若干關(guān)鍵問題研究與應(yīng)用.pdf

1、後多大學(xué)博士學(xué)位論文生物信息數(shù)據(jù)挖掘若干關(guān)鍵問題研究與應(yīng)用院系：專業(yè)：姓名：指導(dǎo)教師：完成日期：復(fù)旦大學(xué)博士學(xué)位論文ABSl∽TIlABSTRACTDatamininghasbeenanactivefieldofstudyformorethanadecade，andanumberofalgorithmsaredevelopedtodealwitllallsortsofdatamiIlingproblemsThesealgorithmse

2、mployavarietyofmethodologies，includingstatistics，artificialintelligence，machinelearning，anddigitalsignalprocessingandSOonInrecentyears，dataminingtechnologiesaresuccessfullyappliedtovariousindustriesandhavedemonstratedits

3、knowledgediscoverypowerThefocusisnowontailoringthedataminingtechniquestospecificapplicationsAseriesofbreakthroughshasbeenmadeinlifesciencerecentlyAsaresultofthesuccessfulexecutionoftheHumanGenomeProjectrHGP)andrelatedadv

4、ancementinmodembiologyahugeamountofdatahasbeenaccumulated，providingastrongdatabaseforuncoveringthesecretoflifeBiologicaldataisrichinvarietyandhigh—throughputwimhighdimensionsInessenceithasheterogeneityandnetworkcharacter

5、sthatarefarbeyondthepoweroftraditionalanalyticalmeansTheanalysisofthesebio—databecomesthebottleneckofbiologicalresearchApplyinginformationtechnologytomolecularbiologygivesariseabrandnewfield，bioinformaticsAsalleffectivew

6、ayoffindingtheneedle(biologicalknowledge)fromahaystack(rawdata)，datamininghasalsobecomeincreasinglyimportantinbioinformaticsstudyAsaresultoftheprogressofthegenomeresearchandtheadventofmanyhighthroughputtechnologies，profo

7、undchangesaretakingplaceintheresearchmethodologyofmodembiosciencesEventraditionallyexperimentbasedsubjectsmoreandmoreutilizebioinformaticstoolstointerpretexperimentaldataandtogatherfruitfidhintsonexperimentaldesignTheref

8、ore，findingeffectiveanalysistoolsbecomeapressingissuefortheadvancementofmodembiologicalsciencesComputerscientistshavealotofinterestsingettinginvolvedinbioinformaticsdataanalysisAsamajormethodologydatamininghasapromisingr

9、oletoplayinthisendeavorHoweverresearchesondatamininginbiologyarestillinapreliminaryphaseandarefacingalotofchallengesHowtoapplyvariousdataminingtechniqueseffectivelytobiologicalinformationanalysisisahotsubjecttodayTllisin

10、cludesfindingdataminingsystemarchitecture，newalgorithms，andnewmethodologyetcthataresuitableforbioinformaticsdataanalysisTllisdissertationfocusesondataminingapplicationinbioinformaticsn地mainresultsofthisstudyaresummarized

11、asfollows：1GeneexpressiondataanalysisAfterreviewingtheexistinggeneexpressiondatamining，analysismodelsongeneexpressionsimilarityexplain，peculiarexpressiongeneanalysis，andbypathanalysismodeletchavebeenproposedEachmodelisma

12、ppedtogivenanalysisflowanddataminingalgorithmsThesemodelsarepracticalforbindataanalysis2BioinformaticsdataminingsystemarchitectureTobuildabetterdataminingapplicationframeworkforbioinformaticsdataanalysis，wedevelopeda4tie