文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 自然科学 > 数学 > A modified rough c-means clustering algorithm based on hybrid imbalanced measure
侵权投诉

A modified rough c-means clustering algorithm based on hybrid imbalanced measure

InternationalJournalofApproximateReasoning55(2014)

A modified rough c-means clustering algorithm based on hybrid imbalanced measure

1805–1818

ContentslistsavailableatScienceDirect

InternationalJournalofApproximateReasoning

A modified rough c-means clustering algorithm based on hybrid imbalanced measure

A modified rough c-means clustering algorithm based on hybrid imbalanced measure

doc.guandang.net

Amodi edroughc-meansclusteringalgorithmbased

onhybridimbalancedmeasureofdistanceanddensity

TengfeiZhanga, ,LongChena,FuminMab

a

bCollegeofAutomation,NanjingUniversityofPostsandTelecommunications,Nanjing,210023,ChinaCollegeofInformationEngineering,NanjingUniversityofFinanceandEconomics,Nanjing,China

articleinfoabstract

Traditionalc-meansclusteringpartitionsagroupofobjectsintoanumberofnon-over-

lappingsets.Roughsetsprovidemore exibleandobjectiverepresentationthanclassical

setswithhardpartitionandfuzzysetswithsubjectivemembershipfunctionforagiven

dataset.Roughc-meansclusteringanditsextensionswereintroducedandsuccessfully

appliedinmanyreallifeapplicationsinrecentyears.Eachclusterisrepresentedbyarea-

sonablepairoflowerandupperapproximations.However,themostavailablealgorithms

paynoattentiontothein uenceoftheimbalancedspatialdistributionwithinacluster.

Thelimitationofthemeaniterativecalculationfunction,withthesameweightforallthe

dataobjectsinalowerorupperapproximation,isanalyzed.Ahybridimbalancedmeasure

ofdistanceanddensityfortheroughc-meansclusteringisde ned,andamodi edrough

c-meansclusteringalgorithmispresentedinthispaper.Toevaluatetheproposedal-

gorithm,ithasbeenappliedtoseveralrealworlddatasetsfromUCI.Thevalidityofthis

algorithmisdemonstratedbytheresultsofcomparativeexperiments.

©2014ElsevierInc.Allrightsreserved.Articlehistory:Received6November2013Receivedinrevisedform4May2014Accepted5May2014Availableonline13May2014Keywords:Roughc-meansclusteringHybridimbalancedmeasureRoughsettheory

1.Introduction

Inthepastdecades,dataclusteringhasbecomeatopicofgreatinteresttoresearchersandhasbeenusedfrequentlyinreallifeapplications[1].Thefundamentalobjectiveofdataclusteringistogroupsimilarobjectsinoneclusteranddividedissimilarobjectsintodifferentclusters.Themajoravailableclusteringtechniquesaregenerallyclassi edintopartitionalclusteringandhierarchicalclustering,basedonthepropertiesofthegeneratedclusters[2].K-means,viz.,c-meansclusteringalgorithm[3],whereeachclusterisrepresentedbythemeanvalueoftheobjectsinthecluster,isprobablystillthemostcommonlyusedpartitioningmethod.Buttheclassichardc-means(HCM)techniqueunambiguouslyassignsanobjectpreciselytoasingleclusterwithcrispboundary.Inpractice,clusterboundariesmaydisplaycharacteristicsofoverlap.Itisnotthebestwaytoabsolutelydividetheboundaryobjectsintoonecluster.Insuchcases,anobjectintheboundaryshouldbelongtomorethanonecluster[4].

Theuseofsoftcomputinginclusteringhasbeenreportedinliterature[5,6].Fuzzysetsprovidesuchmethodsbytheconceptofmembershipfunction,usingalgorithmssuchasfuzzyc-means(FCM)[7,8].Itenablese cienthandling*Correspondingauthor.Tel.:+8615850597135.

E-mailaddress:zhangtf@njupt.edu.cn(T.Zhang).

doc.guandang.net

0888-613X/©2014ElsevierInc.Allrightsreserved.

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email