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 > 天文/地理 > A Clustering Method for Spatio-temporal Data and Its Application to Soccer Game Records
侵权投诉

A Clustering Method for Spatio-temporal Data and Its Application to Soccer Game Records

AClusteringMethodforSpatio-temporalDataandItsApplicationtoSoccerGameRecords

ShojiHiranoandShusakuTsumoto

DepartmentofMedicalInformatics,ShimaneUniversity,SchoolofMedicine,

89-1Enya-cho,Izumo,Shimane693-8501,Japan

hirano@ieee.org,tsumoto@computer.org

Abstract.Thispaperpresentsanovelmethodfor ndinginterestingpatternsfromspatio-temporaldata.First,weperformapairwisecom-parisonofspatio-temporalsequencesusingthemultiscalematching,tak-ingintoaccounttherequirementsformultiscaleobservation.Next,weconstructtheclustersofsequencesusingrough-setbasedclusteringtech-nique.Experimentalresultsonrealsoccergamerecordsdemonstratedthatthemethodcoulddiscoversomeinterestingpasspatternsthatmaybeassociatedwithsuccessfulgoals.

1Introduction

Clusteringofspatio-temporaldataprovidesanew,data-orientedwayofdiscov-eringinterestingknowledgeaboutthemovementoftargets.Ithasbeenreceivingmuchattentioninvarious eldssuchascrimeresearch,meteorologyandsportsdataanalysis,asatoolforrevealingcommonbehavioralcharacteristicsofthetargets.Forexample,insoccergame,byclusteringspatio-temporaldataaboutthepasssequence,onemayobtaininterestingknowledgeaboutthestrategyofateamlike’frequentuseofright-sideattack’.

Thispaperpresentsanovelmethodforvisualizinginterestingpatternshiddeninthespatio-temporaldata.Asatangibledataweemploysoccergamerecords,astheyinvolvethemostimportantproblemsinspatio-temporaldatamining–thetemporalirregularityofdatapoints.Especially,wefocusondiscoveringthefeaturesofpasstransactions,whichresultedinsuccessfulgoals,andrepresentingthedi erenceofstrategiesofateambythepassstrategies.

Therearetwopointsthatshouldbetechnicallysolved.First,thelengthofasequence,numberofdatapointsconstitutingasequence,andintervalsbetweendatapointsinasequenceareallirregular.Apasssequenceisformedbycon-catenatingcontiguouspassevents;sincethedistanceofeachpass,thenumberofplayerstranslatingthecontiguouspassesarebynaturedi erence,thedatashouldbetreatedasirregularsampledtime-seriesdata.Second,multiscaleob-servationandcomparisonofpasssequencesarerequired.Thisisbecauseapasssequencerepresentsbothglobalandlocalstrategiesofateam.Forexample,asaglobalstrategy,ateammayfrequentlyuseside-attacksthancounter-attacks.Asalocalstrategy,theteammayfrequentlyuseone-twopass.Bothlevelsof´ezaketal.(Eds.):RSFDGrC2005,LNAI3641,pp.612–621,2005.D.Sl c Springer-VerlagBerlinHeidelberg2005

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