Study The Application Of Data Mining Technology In Prediction Of Customer Churn In Insurance Company

The basic idea of Data Mining is that use the data through the establishment of mathematical model of the method to find hidden business rules. With data mining technology development, the importance of data mining has been recognized more and more people, it is known to Many industries in foreign countries has been a successful application. For example, the insurance industry applications are mainly customer relationship management, customer fraud analysis, customer churn analysis, customer consumption patterns, analysis, marketing analysis. In the country with the importance of data mining, data mining techniques applied research are increasingly broad, in which the insurance industry, customer churn analysis is a hot topic. Customer churn analysis is the loss of customers through the past, historical data analysis to identify the characteristics of the user may surrender promptly take appropriate measures to reduce customer churn happening. This enterprises to reduce operational costs and improve business performance has extremely important significance.This paper begins by describing the definition of data mining, function and processes, analysis of the content as well as the CRM system framework, and points out the reasons for the loss of insurance company clients and the achievement of customer churn prediction of the need for insurance companies, insurance companies will be on this basis, classify the reasons for customer churn analysis to be insurance companies lose customers KPI and the corresponding countermeasures. Second, the focus on the design of the insurance company's customer churn prediction models. Decision tree for classification mining algorithms commonly used in the analysis, pointing out the existence of the issue, based on the weighted attributes and pre-pruning strategy for improved decision tree classification algorithm for mining. The algorithm can be a better solution to the insurance industry in data mining large data volume and high efficiency requirements. At the same time improved algorithm based on decision tree mining on churn prediction models in the framework of the overall design of the customer-related data collection, integration, under the proposed restructuring of KPI data. And a clear customer churn prediction model built under the premise of thought is given to improve the algorithm based on decision tree mining customer churn prediction model building process in detail. Then, specific commercial life insurance company, will this set up customer churn prediction model application, a detailed analysis of the model application process involved in the data cleansing, data conversion as well as through improved decision tree classification algorithm for mining customer churn prediction model established.Finally, prediction models to assess the results and analysis are given specific policy proposals, effectively improve customer turnover, is also shown that decision tree classification algorithm to improve the effectiveness and practicality.