電信企業(yè)客戶流失預(yù)測(cè)模型研究

1、 Subject : Research on Prediction Model of Telecom Customers Churn Engineering Field : Computer technology Name : Zhong Ji (Signature) Instructor : Li Zhanli

2、 (Signature) ABSTRACT With the deepening reform of telecommunication systems, the competition between operators of domestic telecommunications industry becomes increasingly fierce. Meanwhile, operators

3、 tend to follow a variety of promotional activities and an endless stream of advertising to attract new customers in order to gain more customers and take up a larger share of the market. However, statistics found that t

4、he development of a new customer costs 7 times higher than to keep an old customer, and if “user retention rates” increase 5%, 85% profit growth is expected to bring to operators. So old customers retained is directly re

5、lated to the interests of operators, both customer loss and the loss of traffic volume will have a profound impact on operators. For this problem, the paper studies characteristics of losing customers with data mining, a

6、nd thus to predict the loss and assess the consequences of the loss, customer retention measures taken to prevent leading management crisis, and also to enhance the competitiveness of telecommunications companies. The ma

7、in fields of application research of data mining in the customer relationship management in the telecommunication industry and the related application are studied in this paper, and the causes of customer churn are also

8、analyzed in order to differentiate between losing customers, thus leading to wastage of different standards. There prediction algorithms are decision trees, neural networks, support vector machines and logistic regressio

9、n, etc. but all of them have advantages and disadvantages. C5.0 decision tree, C&T decision trees, neural networks and support vector machines are used in this paper and results show that C5.0 decision tree has the h

10、ighest resolution accuracy rate of 90%, followed by SVM reached more than 80%. In order to increase the models’ feasibility and accuracy of prediction for different data, a comprehensive model is put forward, which is ac

11、hieved by adding a confidence interval above the models for integrated. Finally, the integrated model is built. This proposed model combines the advantages of a variety of forecasting methods, improving the accuracy of f

12、orecasts and reducing the risk prediction, and it provides a theoretical basis for loss 目 錄 I 目 錄 1 緒論 ..................................................................................................................

13、...................... 1 1.1 選題的背景和意義 ........................................................................................................... 1 1.2 國(guó)內(nèi)外研究現(xiàn)狀分析 .........................................................

14、............................................. 2 1.3 主要研究?jī)?nèi)容、目標(biāo)和方法 .......................................................................................... 3 1.4 論文的結(jié)構(gòu) ....................................................

15、.................................................................. 4 2 客戶流失分析方法 ................................................................................................................. 5 2.1 客戶流失分析相關(guān)理論 ........

16、.......................................................................................... 5 2.1.1 數(shù)據(jù)挖掘理論的產(chǎn)生與發(fā)展 .................................................................................. 5 2.1.2 數(shù)據(jù)挖掘的任務(wù) .........

17、............................................................................................. 7 2.1.3 數(shù)據(jù)挖掘系統(tǒng)組成 .................................................................................................. 8 2.1.4 數(shù)據(jù)

18、挖掘過(guò)程 .......................................................................................................... 8 2.1.5 數(shù)據(jù)挖掘的常用技術(shù) ..........................................................................................

19、.... 9 2.1.6 數(shù)據(jù)挖掘的類型 .................................................................................................... 12 2.2 數(shù)據(jù)挖掘技術(shù)在電信行業(yè)應(yīng)用 .............................................................................

20、....... 13 2.3 客戶流失及其分析方法 ................................................................................................ 15 2.3.電信客戶流失的類型 ................................................................................

21、............... 15 2.3.2 客戶流失分析方法 ................................................................................................ 16 2.4 客戶流失分析常用算法 .......................................................................

22、......................... 17 2.4.1 支持向量機(jī)算法原理 ............................................................................................ 18 2.4.2 神經(jīng)網(wǎng)絡(luò)算法原理 ................................................................

23、................................ 19 2.4.3 決策樹(shù)算法原理 .................................................................................................... 21 2.4.4 Logistic 回歸二元分類原理 ..........................................

24、........................................ 22 3 客戶流失組合預(yù)測(cè)模型 ....................................................................................................... 23 3.1 數(shù)據(jù)挖掘軟件系統(tǒng) ..........................................

25、.............................................................. 23 3.2 SPSS Clementine 12.0 數(shù)據(jù)挖掘系統(tǒng) ........................................................................... 23 3.3 SPSS Clementine 12.0 的基本操作方法 ........

26、............................................................... 25 3.4 數(shù)據(jù)流的基本操作方法 ................................................................................................ 27 3.5 Clementine 12.0 的主要模型 ...........

27、.............................................................................. 28 4 電信企業(yè)客戶流失預(yù)測(cè)模型應(yīng)用 ....................................................................................... 29 4.1 商業(yè)理解 ....................

28、.................................................................................................... 29 4.2 數(shù)據(jù)理解 ...........................................................................................................