資源描述：
《數(shù)據(jù)挖掘技術(shù)在CDN網(wǎng)絡(luò)中的應(yīng)用研究》由會員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。
1、北京郵電大學(xué)碩士學(xué)位論文數(shù)據(jù)挖掘技術(shù)在CDN網(wǎng)絡(luò)中的應(yīng)用研究姓名：馬玉會申請學(xué)位級別：碩士專業(yè)：應(yīng)用數(shù)學(xué)指導(dǎo)教師：阮傳概20090215北京郵電大學(xué)碩士學(xué)位論文由技術(shù)兩者結(jié)合的可能性。3．研究了數(shù)據(jù)挖掘技術(shù)，特別是Web數(shù)據(jù)挖掘技術(shù)，并總結(jié)出Web數(shù)據(jù)挖掘技術(shù)發(fā)展方向。4．研究了數(shù)據(jù)挖掘技術(shù)中的聚類算法在用戶集群劃分中的應(yīng)用，并論述了客戶細(xì)分的基本理論、方法和步驟。CDN的應(yīng)用越來越廣泛，并且解決網(wǎng)絡(luò)帶寬不足的問題。滿足人們各種需求的CDN也必將迎來更好的發(fā)展機(jī)遇，因此這種把數(shù)據(jù)挖掘技術(shù)和CDN內(nèi)容路由技術(shù)相結(jié)合，直接運(yùn)
2、用于基于全局負(fù)載均衡和基于客戶端探測技術(shù)相結(jié)合的設(shè)計模式也必將更好的體現(xiàn)它的實用價值和參考價值。關(guān)鍵詞：路由優(yōu)化全局負(fù)載均衡基于用戶端探測用戶集群數(shù)據(jù)挖掘聚類算法Ⅱ北京郵電大學(xué)碩士學(xué)位論文DATAMININGTECHNoLoGYINCDNAPPLIEDRESEARCHABSTRACTWiththedevelopmentofcomputernetworktechnology,networkhasinfiltratedintoeverycomeroftheworld．However,whenweontheInternet，w
3、eoftenencountertoslowresponses，andevenerrorresponses．Thedistancebetweentheactualperformanceofthenetworkandthequalityofserviceexpectationsisverylarge，andithasbecomethefocus．Inviewofthis，establishingacomprehensivemiddlelayerCDN(ContentDeliveryNetwork,ContentDistrib
4、utionNetwork)intheexistingnetworksisthemostpopularselectioncurrently．Betweenthis，thecontentroutingsystemistheessentialpartoftheCDNarchitecture．Because，accordingtotheuser’srequestandtheconditionoftheedgeserver,thebestedgeserverisselected．Andinthedesignofnewcontent
5、routingsystem,theUSerclusterdivisionisalsoveryimportant．Anddataminingistheprocedureofextractingofimplicit，original，usefulknowledgeinthedatabase，whichisalreadyappliedinmanyfieldsinrecentyears．Clusteringanalysisisoneofthemaintechnologymeasuresintheresearchondatamin
6、ingwithamassoftheoriesandmethodsachieved．WiththeestablishmentofDecisionSupportSystemsuchasdatawarehouseandtherequirementofbusinessintelligenceinthedataintensiveenterprises，datamininghasbeenusedinmanynewapplicationsandtheresearchondataclusteringisfacedwithalotofne
7、wchallenges．CDNisoneofthetypicaldataintensivewebsite，dividingthecustomersandsupplyingdifferentedgeserverstodifferentuser’SclustershasalreadybecomeinstantdemandtoimproveCDN’Sperformance．Anewalgorithmclusteringthedatasetswithmixednumericaland北京郵電大學(xué)碩士學(xué)位論文categorical
8、valuesiSresearchedbasedontherequirementofuserssegmentationandthecharacteristicofthedataintheCDNapplicationisproposedinthepaper．Ontheconditionofnotusethird-part
