資源描述:
《數(shù)據(jù)挖掘在電信客戶流失中的應用》由會員上傳分享,免費在線閱讀,更多相關內容在學術論文-天天文庫。
1、哈爾濱工業(yè)大學管理學碩士學位論文摘要隨著數(shù)據(jù)挖掘技術的不斷發(fā)展,人們越來越關心它在各個領域的應用,它是利用已有的歷史數(shù)據(jù)通過建立模型的方法找出隱含的業(yè)務規(guī)則。電信行業(yè)是典型的數(shù)據(jù)密集行業(yè),而且目前電信運營商們都在經(jīng)歷著不同程度的客戶流失,所以運用數(shù)據(jù)挖掘技術找出電信流失客戶的特征顯得尤為重要。本文的研究目的就是應用數(shù)據(jù)挖掘技術發(fā)現(xiàn)流失客戶的特征,建立流失預測模型,應用模型預測流失客戶名單,針對流失客戶制定有效的挽留策略。本文的主要工作包括以下幾個方面:首先從業(yè)務角度,對電信行業(yè)的業(yè)務流程和電信客戶流失原因進行分析,將客戶流失原因和電信行業(yè)的業(yè)務系統(tǒng)相接合,總結出與客戶流失相關的數(shù)據(jù)
2、資料,完成從商業(yè)角度向數(shù)據(jù)角度的轉換。然后從技術角度,總結了數(shù)據(jù)挖掘技術的理論與方法。詳細描述了神經(jīng)網(wǎng)絡挖掘算法和決策樹算法的原理及過程,為后面章節(jié)的實際建模打下鋪墊。接著在前面的基礎上又介紹了電信行業(yè)客戶流失模型建立的原則,探討了數(shù)據(jù)挖掘過程的標準化模型。詳細描述了CRISP-DM標準數(shù)據(jù)挖掘過程的6個階段,并結合電信行業(yè)自身的特點,論述了商業(yè)理解、數(shù)據(jù)理解、數(shù)據(jù)準備、建立模型、模型評估、模型應用這六個階段的工作。最后又詳細探討了流失客戶的挽留價值,詳細描述了電信客戶生命周期各個階段的特點,分析客戶流失的原因,針對流失客戶制定有效的客戶挽留策略。關鍵詞數(shù)據(jù)挖掘;客戶流失;跨行業(yè)的
3、標準數(shù)據(jù)挖掘流程;分類回歸樹;挽留價值-I-哈爾濱工業(yè)大學管理學碩士學位論文AbstractWiththedevelopmentofdataminingtechnology,peoplebecomemoreandmoreinterestedinitsapplicationinvariousfields.Dataminingistofindoutthehiddenbusinessrulesthroughtheestablishmentofthemodelbyusinghistoricaldata.Thetelecomindustryisatypicaldata-intensivein
4、dustry,andcurrentlytelecomoperatorsaresufferingdifferentdegreesofcustomerschurn.Therefore,itisparticularlyimportanttousedataminingtoidentifythecharacteristicsofthechurncustomer.Thepurposeofthispaperistodetectthecharacteristicsofthelostcustomer,establishchurnpredictionmodel,establishthelostcust
5、omerlistbyusingthemodel,andmakesaneffectivedetainmentstrategyagainstthechurncustomers.Themaincontentsofthispaperincludethefollowingaspects:Firstly,fromthebusinessview,thispaperanalysesthebusinessprocessoftelecomindustryandthereasonofcustomerchurn.Thenthispapercombinesthereasonofcustomerchurnan
6、dthetelecombusinesssystem,achievingtheconversionfromcommerceviewtodataview.Secondly,fromthetechnicalview,thispapersummarizesthetheoryandmethodofdatamininganddescribesthealgorithmofneuralnetworkanddecisiontreeindetail.Then,thispaperdescribestheprinciplesofestablishingthemodelofthetelecomindustr
7、ycustomerchurn,anddiscussesthestandarddataminingprocess.ThispaperdescribesthesixstagesofCRISP-DMstandarddataminingprocess,anddiscussesthebusinessunderstanding,dataunderstanding,datapreparation,modelbuilding,modelevaluation,andmodelimple