欢迎来到天天文库
浏览记录
ID:33312878
大小:545.70 KB
页数:67页
时间:2019-02-24
《硕士论文关联则的挖掘及其在入侵检测中的应用精选》由会员上传分享,免费在线阅读,更多相关内容在应用文档-天天文库。
1、西南交通大学硕士学位论文关联规则挖掘及其在入侵检测中的应用姓名:荣文亮申请学位级别:硕士专业:计算机应用技术指导教师:杨燕20080124西南交通大学硕士研究生学位论文第Ⅱ页Abstract.Withtherapid。developmentofinternet,theprocessingdatainmanyapplicationareasisintheformofdatastream,suchasthemonitoringinformationfromnetworkmonitoringsystemwhichisakind
2、ofdatastream.Thesedatastreamscontainagreatdealofusefulinformationandknowledge,andabnormitiesontheinternetareeffectivelydiscoveredbyminingdatastream,whichareusedforintrusiondetection.Dataminingisamethodwhichobtainsusefulknowledgeandinformationfromlargeamountsofdat
3、as.Associationrulesminingisanimportantmethodindatamining.Initially,theresearchesonassociationru.1esalgorithmmainlyfocusedonthetraditionaldatamininginthepa§tfewyears,withtheconceptofdatastreamproposed,theresearchersbeginpayingattentiontothedatastreammining.However
4、,becausethedataoverdatastreamiscontinuous,rapid,andinfinite,SOthetraditionaldataminingtechnologydon’tperformwellforthedatastream.Therefore,agrowingnumberofresearchersstartanewresearchonassociationrulesminingoverdatastream.Thispapersystematicallyintroducestraditio
5、nalassociationrulesminingalgorithmsandtheirkeytechnology,andproposesanimprovedalgorithmaccordingtotheirkeytechnology;Inaddition,thepaperalsoanalyzesthecharacteristicsofthedatastreamindetail,andmainlyintroducestWOclassicalalgorithmsonfrequentpatternminingoverdatas
6、treamatthesametime..Finally,itputsforwardafrequentclosedpatternsminingalgorithmoverdatastreamwiththeclosedpatternsminingtechnologywhichisoneofmethodsoftheassociationrulemining.・’Finally,thispaperappliesthetwoalgorithmsproposedabovetotheanomaly—basedintrusiondetec
7、tion,anddesignsanintrusjondetectionII西南交通大学硕士研究生学位论文第1II页systemmodelbasedontheassociationrulesmining.Themodelusesanomalydetectiontechnology,whichdividestheprocessesofin.trusiondetectionintotwoparts:detectionprocessandlearningprocess.Thefirstprocessperformswiththe
8、algorithmofassociationrulesmini.ngoverdatastream,andthesecondprocessperformswithtraditionalassociationminingalgorithm.Toacertainextent,themodelcanimprovereal.t
此文档下载收益归作者所有