Research And Application Of The Telecom Customer Churn Prediction Based On Data Mining

With the rapid development of data mining technology,data mining has been more and more recognized and applied to many research fields.In the study of telecom customer churn prediction,data mining is based on historical data and current data mining algorithm as the foundation,find one or more data,establish corresponding method,and explore the relationship between data,make the corresponding prediction according to data mining,so as to provide constructive suggestions for the operation of decision makers,retain the potential loss of customers,and allows operators to benefit more.This paper mainly uses the existing data mining technology to analyze and forecast the telecom customers,combined with the combination of the theory and methods of the combination model to achieve the goal of improving the accuracy of prediction.This paper chooses decision tree,neural network,SVM and Logistic Regression,four classic prediction methods,compares and analyzes various prediction accuracy using SPSS Clementine,and then uses the weighted method to combine the models,finally achieves the purpose of improving the prediction accuracy.The combined model integrates the advantages of several models,the results show that the accuracy of the prediction results is improved,and the reliability is also improved.The main innovations of this paper are as follows: in order to more accurately reflect the user's features,this paper uses rough set algorithm to screen multiple attributes of users.Considering the defects of single prediction model,this paper combines the decision tree,SVM,neural network,and Logistic regression into a comprehensive prediction model linearly,using the method of AHP,so as to improve the overall prediction accuracy.The experiment results show that the precision of the combined prediction model is much larger than that of the single prediction model and other combined prediction models.