基于遺傳算法的快速響應(yīng)多關(guān)聯(lián)數(shù)據(jù)挖掘.pdf

1、上海大學(xué)碩士學(xué)位論文基于遺傳算法的快速響應(yīng)多關(guān)聯(lián)數(shù)據(jù)挖掘姓名：竇文翔申請學(xué)位級別：碩士專業(yè)：軟件與理論指導(dǎo)教師：吳耿鋒20080701上海大學(xué)碩士學(xué)位論文ABSTRACTWiththedevelopmentofstoringtechnique，moreandmoredataarestoredasvariouscomplicatedrelationalstructureonrelationaldatabaseHoweverexisting

2、dataminingtechniquesCallnotdealwiththesemulti—relationaldataAsaresult，themaintaskoftoday’SdataminingresearchfieldishowtofindthevaluableinformationquicklyandeffectivelyfrommultirelationaldataInthispaperweanalyzetheloweffi

3、cientproblemappearinginthemostmethodswhenminingrulesindatanaplescontaininglongitemsetsandtheproblemhowtoprocessmultirelationaldataForthelowefficientproblem，weproposeallefficientminingmodelformakingquickresponsetousersand

4、providingthemwithneededassociationrulesThismodelusesmodifiedgeneticalgorithmtomillemaximalfrequentitemsets，andthenletusersselectsomeinteresteditemstodeducethefinalassociationrulesSo，itjustscansthedatabasefortheneededrule

5、sForprocessingmultirelationaldataweproposeadistributedminingmethodwhichminesrulesinmultiplerelationaltablesatthesametimeThismethod，firstminesantecedentrulesfromthetablecontainingchiefattributes，andthenminesmultipleconseq

6、uentrulesfromotherrelationaltablescontainingaffiliatedattributesaccordingtothef0rei印IDWeuseanewextendedassociationruletoexpresstheminedknowledgeInadditionaquickresponsemultirelationaldataminingsystemisconstructedthroughc

7、ombiningtheaboveminingmodelandthedistributedminingideaTherefore，thissystemownsalladvantagesItcanmakequickresponsetousers，provideneededrules，processmultirelationaldataandshowtheextendedassociationruleshavingfinepatternand