资源描述：
《毕业论文外文翻译-数据挖掘—聚类分析》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、电气信息工程学院外文翻译英文名称：Datamining-clustering译文名称：数据挖掘—聚类分析专业：自动化姓名：****班级学号：****指导教师：******译文出处：Datamining：IanH.Witten,EibeFrank著Clustering5.1INTRODUCTIONClusteringissimilartoclassificationinthatdataaregrouped.However,unlikeclassification,thegroupsarenotpredefined.Instead,thegroupingisaccomplishedb
2、yfindingsimilaritiesbetweendataaccordingtocharacteristicsfoundintheactualdata.Thegroupsarecalledclusters.Someauthorsviewclusteringasaspecialtypeofclassification.Inthistext,however,wefollowamoreconventionalviewinthatthetwoaredifferent.Manydefinitionsforclustershavebeenproposed:lSetoflikeelemen
3、ts.Elementsfromdifferentclustersarenotalike.lThedistancebetweenpointsinaclusterislessthanthedistancebetweenapointintheclusterandanypointoutsideit.Atermsimilartoclusteringisdatabasesegmentation,whereliketuple(record)inadatabasearegroupedtogether.Thisisdonetopartitionorsegmentthedatabaseintocom
4、ponentsthatthengivetheuseramoregeneralviewofthedata.Inthiscasetext,wedonotdifferentiatebetweensegmentationandclustering.AsimpleexampleofclusteringisfoundinExample5.1.Thisexampleillustratesthefactthatthatdetermininghowtodotheclusteringisnotstraightforward.AsillustratedinFigure5.1,agivensetofda
5、tamaybeclusteredondifferentattributes.Hereagroupofhomesinageographicareaisshown.Thefirstfloortypeofclusteringisbasedonthelocationofthehome.Homesthataregeographicallyclosetoeachotherareclusteredtogether.Inthesecondclustering,homesaregroupedbasedonthesizeofthehouse.Clusteringhasbeenusedinmanyap
6、plicationdomains,includingbiology,medicine,anthropology,marketing,andeconomics.Clusteringapplicationsincludeplantandanimalclassification,diseaseclassification,imageprocessing,patternrecognition,anddocumentretrieval.Oneofthefirstdomainsinwhichclusteringwasusedwasbiologicaltaxonomy.Recentusesin
7、cludeexaminingWeblogdatatodetectusagepatterns.Whenclusteringisappliedtoareal-worlddatabase,manyinterestingproblemsoccur:lOutlierhandlingisdifficult.Heretheelementsdonotnaturallyfallintoanycluster.Theycanbeviewedassolitaryclusters.However,ifac
