数据挖掘技术及应用研究

《数据挖掘技术及应用研究》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、国内图书分类号{TP315苗脚m船，。善792032单位代码：100如学号；密级：S200218013公开北京工业大学工程硕士学位论文题目．．熬掘挖掘拉丕臣应目的班究一一英文并列题目圈E跚A衄巡驰AIAM圜毯G蚀璐创哩U￡A]隗巡研究生姓名：宝攫研究方向：堑搔控翅职称：直级王捏娅职称：高级王程哑论文报告提交日期2嫂』生5且学位授予日期授予单位名称和地址j丘童王业去堂j丘毫壶掘田区垩基园．1盟置￡l鲤蝗2^UstIaCLAbstractAtthecurrentinformationage，theresearchofdataminin

2、gtechnologyisgettingincreasinglyhot，becauseofthehugevolumesofdata,andtheapplicationofinformationextractedfromverylargedatabase．aswellastheimpactofinformationovertheenterprisestrategicdecisionprocess．Atpresent，themainresearchfieldofdataminingincludesdatasummarizing，ass

3、ociationrules，classifyinlg，clustering，etc⋯Takingtheapplicationofdataminingtechnologiesbasedondatawarehouse硒themainconcentrationofthepaper，lgetthismainlybytheassociationrulesandclusteringtechnologyresearch．Theadvancedalgorithmsofapplyingthetwomethodstothedevelopedproje

4、ct黜discussedwithinthispaper．Themainresearchachievementsandcontentsofthethesisinclude：1．TheresearchofdatawarehouseandOLAFtechnique．Thedatawarehouseisprovided、Ⅳimthedataplatformfordataminingsystem，andhenceisfirstlybuiltupbeforeimplementofdatamining．OLAPisbasedOilthedata

5、warehouse，whichisaverifyingandanalysist001．DataminingandOLAFCanrectifyinfunctioneachother，thattheyassisttheenterprisestrategicdecisionprocess．Datawarehouse,dataminingand0LAP咖besufficientlyintegrated．thatiSOnLineAnalysisMining(OLAM)．2．Theimprovementonthetypicalassociat

6、ionruleminingalgofithm．ThepaperintroducesthetypicalApdodalgorithm，andanalyzesitsdrawbackthatithavetostartfromscratchoncalculatingtheLargeItemsotsanddiscoveringassociationruleswhenminimumsupportandminimumconfidence，aswellastransactiondatabase，arechanged,andthereforepre

7、sentsaimprovedalgorithm：quickerupdatingalgorithmforassociationrulemining，basedonApfiorialgorithm．TheresultshowsthattheperformanceoftheimprovedalgorithmissuperiortothatofApriorialgorithm．3．Theoptimizationoftheclusteringalgorithm．Thethesismainlyanalyzesclusteringalgorit

8、hmbasedOntheclusteringanalysistechnologyindataminingarea,Besides，thepaperdetailedlyintroducedtheK—Meansclusteringalgorithm，a