资源描述：
《聚类分析与朴素贝叶斯分类在客户价值预测中的应用研究》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、合肥工业大学硕士学位论文聚类分析与朴素贝叶斯分类在客户价值预测中的应用研究姓名：赵敏申请学位级别：硕士专业：管理科学与工程指导教师：毛雪岷;倪志伟20100401ClusteringandNa^veBayesiana1gorithmincustomervalueforecastingAbstractInrcccntycars,marketinghasbccomcmorccompctitivc,thcvariousbusinessesexcelnotonlyinproductquaIity,buta1soinbus
2、inessmanagementandmarketingofinformationtechnologyhasalduct—centrictoacustomer—centrisomadegreatimprovementsbegantoshiftfromaprogy."Wimtheincrnumberofcudepartment,easeinmerstomers.ttheimportanceoferersimu1taneouintainingcuanIforbusinomersonlyacchandisesales,g
3、enehesa1estecustomersaredlargejee,duetodjfflevelsfordiffent,ofmaportcustent,「entnumberoftre1atewdemands.Wi}ththedeve1opmentofinformainformationtechnologybegf為rietqet1yenhasrsthebussbyspendneepretcustomeuetoputforwardnhno1ogy,dataminingusedincorporateppeadatap
4、rocichmakesenterprK—meansc1usteringaIg0r•thmtostudytheimportanee0feustomerissuestothedatawarehouscforcus10merdaIaobjecIs,iryingiobu•11donIhecurrcnIdaava1ueofIhemode1,andmode1aPproPriateana1ysis,t0identifyfact0rsthatpredicted谢haVarietyofva1uab1einformat•1onish
5、iddent0he1PimPr0veenterPrisecust0merre1ti0nsh■1Pmanagcmcnt9enterpr•1sedeVe10pmentstrategiestofurtherProvidethebassf0rrcfcrcncc,s0thqtenterprisesCanused1rferentmarketingstrategiesf0rdiffeentc.uistomersAAndth■1swi11Pr0videabasist.orbusincssdecisionS・Requireente
6、rprisestoprovisificatanc1asngdataminingtechniquesionusia1rdrsthegreatestvalu1ytoince.ThisisnotonurcustometcaThisNa'fveBayesicombinedwiththcmarkctci1v1t1esasethecus1ueinthetomernumberofeustomers,buta1sotoincreasethecusinhitirethenumberofgh—va1uecustomers.erpri
7、^sevaluetobeighforthoseciInordertocarryoutmoretargeHl1atthe11ng,.e叮j1at1ve1yh111impQrtarignitoesionshipmay.NaiveBaion,anditgmustomersbetterseenttomegemeyehatheoryfoundhillilac1cct•0n•sVcry•mp0rtantf0ra0nd•t•0na1ndependenceasSumpt0n■EXperhrmtcaQ15geugarreaen0F
8、eaecacuracyandcanta1cust（）mers■e:y'w0rdS•■DataM■1n•ngc1uc・lcecatecuNB—•basedc1ass•f•crwh•chUScStenta1resu1tssh0wthatthea1g0rPredctm0reh■曲Va1uep0ter■1ngNa/fVeBayesofc1ficrattracyeassaion,c
