基于遺傳算法的數(shù)據(jù)庫(kù)知識(shí)發(fā)現(xiàn)的過(guò)程與算法的研究.pdf

1、中國(guó)科學(xué)技術(shù)大學(xué)碩士學(xué)位論文基于遺傳算法的數(shù)據(jù)庫(kù)知識(shí)發(fā)現(xiàn)的過(guò)程與算法的研究姓名：徐楊申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：控制理論與控制工程指導(dǎo)教師：王俊普2001.5.1里壁蘭墊查盔蘭堡主鎏塞一——』垡翌墜!∑ABSTRACTKnowledgeDiscoveryinDatabase(KDD)，asallinterdisciplinaryfieldbetweenArtificialIntel[igenceespeciallyMachineLearni

2、ng，andDatabaseTechnologyisahotresearchsubjectincomputerscienceThispaperputsforwardaKnowledgeDiscoveryinDatabaseBasedonGeneticAlgorithm(KDDGA)methodwhichcombinesGeneticAlgorithm(GA)asuccessfullyusedtechniqueofMachineLearn

3、ing，withKDDtechnologyanddoessomefeintireresearchtoltsknowledgediscoveryprocessandapplicationtechnologyInChapter1，lintroduceKnowledgeDiscoveryinDatabaseasanewandpromisingfieldfromtheviewwhy1choosethesetropicTheintroductio

4、nincludesKDD’sproductionanddevelopmentfortheneedsofusers，it’Sdefinitionandtransactionprocesses，moreoverIgiveabriefdescriptionofDataMining(DM)thekeystepofKDDFinallyImentionthemajorresearchworklhavedoneinmythesisChapter2gi

5、vesexplanationtothebasictheoriesofMachineLearningtechnologyespeciallyGAwhichareusedinourresearchofKDDGAThischaptercallbeviewedasthefoundationtheoriesofthispaperChapter3andChapter4viewKDDGAasamultiplytransactionprocessesm

6、odelandhavedonesomeresearchonitChapter3smajortaskistesolvetheproblemhowtotransactorigineddatabeforetheDMsteptomeettheDM’sneedsofUSabledataIntlefirstsectionIintroducenOllllalmethodsusedindatatransactionaccordingtodatatran

7、saction’StaskandsomeofdatatransactionmethodswhichIhaveimprovedarealsogiventhereThesecondsectiondiscusseshowthenewlydevelopedtechnologyDataWarehouseandOn_lineTransactionProcesstransactdatafromtheviewofdatastorageandaccess

8、ionChapter4includesthemajorworkofmyresearchAccordingtothemultiplytransactionprocessesmodelofKDD，IgivethemodeIofKDDGAsystemAtthesalRetimethreealgorithms：ClassificationRulesMiningAlgorithmfromDecisionTreeBaseonGeneticAlgor

9、ithm；WebpagesRecognizingAlgorithmBasedonGeneticAlgorithmandMultiplyAssociationRulesMiningAlgorithmBasedonGeneticAlgorithmwhichareappliedfromthreedifierentKDD’sapplicationfieldsinthekeystepofKDDGAsystemmedelDataMiningalgo

10、rithmBasedonGeneticAlgorithmarebroughtforwardOnthebaseoftheresearchdiscussedinChapter3and4Chapter5usesthealgorithmsdiscussedinChapter4toexperimentwithcorrespondingdatabaseItsresultsarecomparedwithcorrespondresultswhichco

11、meoutfromothermachinelearningalgorithmthatcarlalsealizethesamefunctiontoproveouralgorithm’SfeasibilityandefficiencyKey，Words：KnowledgeDiscoveryinDatabase(KDD)，DataMining，GeneticAlgorithm，MachineLearning，DecisionTree，Clas