cluster analysis

cluster analysis

ID:7323544

大小:826.10 KB

页数:52页

时间:2018-02-11

cluster analysis_第1页
cluster analysis_第2页
cluster analysis_第3页
cluster analysis_第4页
cluster analysis_第5页
资源描述:

《cluster analysis》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库

1、9ClusteranalysisClustering,alsoknownassegmentation,whichIwilldescribemorefullybelow,isthemostwidespreaddescriptivemethodofdataanalysisanddatamining.Itisusedwhenthereisalargevolumeofdataandtheaimistofindhomogeneoussubsets,whichcanbeprocessedandanalysedindifferentways.Sucharequirementispresentinaw

2、iderangeofcontexts,especiallyinthesocialsciences,medicineandmarketing,wherehumanfactorsmeanthatthedataarenumerousanddifficulttounderstand.Inclustering,unlikeclassification(whichwillbedealtwithinalaterchapter),thereisnoparticulardependentvariable,anditishardertocomparetwoformsofclusteringobjective

3、ly.Thereisnoprivilegedcriterionforcomparingtwoalgorithms,suchaswefindintherateofcorrectclassification,andmanycompetingalgorithmsvieforourattention.Onlythemainoneswillbedescribedhere,butitisworthnotingthatthetheoryisconstantlyandrapidlyevolvinginthisfield,especiallynowthattextminingandwebminingareb

4、eginningtogiverisetoproblemsofdocumentclustering.9.1DefinitionofclusteringClusteringisthestatisticaloperationofgroupingobjects(individualsorvariables)intoalimitednumberofgroupsknownasclusters(orsegments),whichhavetwoproperties.Ontheonehand,theyarenotdefinedinadvancebytheanalyst,butarediscovereddu

5、ringtheoperation,unliketheclassesusedinclassification.Ontheotherhand,theclustersarecombinationsofobjectshavingsimilarcharacteristics,whichareseparatedfromobjectshavingdifferentcharacteristics(resultingininternalhomogeneityandexternalheteroge-neity).Thiscanbemeasuredbycriteriasuchasthebetween-clu

6、stersumofsquares(seebelow).Aswithclassification,theessenceofclusteringisthedistributionofobjectsintogroups.However,thisdistributionisnotcarriedoutonthebasisofapredefinedcriterion,andisnotintendedtocombinetheobjectshavingthesamevalueforsuchacriterion.Inotherwords,theclustertowhicheachobjectbelongs

7、isnotknowninadvance,incontrasttotheclassificationprocess.Eventhenumberofclustersisnotalwaysfixedinadvance.ThisisbecausethereisnoDataMiningandStatisticsforDecisionMaking,FirstEdition.Ste´phaneTuffe´ry.Ó2011JohnWiley&Sons,Ltd.Publishe

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。