Research On Commercial Bank's Customer Churn Based On Data Mining

With the implementation of economic globalization and e-commerce, the commercial banks will face more intensive competition, especially in the light of the globalization of competition, as well as resources for high-value customers in the increasingly fierce competition, allowing customers avated disturbance, serious loss of customers, customer acquisition costs increase in bank operating risks increase competitiveness by emendous impact. Study found that in the banking industry, customer retention is the key to a successful CRM strategy, and only when the client is maintained over time, the strategy is profitable, and the success of customers to reduce banks seek new, and potentially risk customers demand, and banks will focus on the establishment of relations and meet the needs of existing customers. Therefore, the face of the current market conditions, commercial banks are required in the development of new customers at the same time, to proceed with customers research. Commercial banks to maintain existing customers can increase the convergence of funds, while saving induced customers to enter the necessary advertising and on cost, thus generating more cash flow and profits. Commercial bank customers and maintain the success of the bank customers rely mainly on the analysis and evaluation of information, primarily rely on data mining technology, as great an unprecedented level of individual customer data in a database makes commercial banks become large and complex - how to effectively deal with , analysis and interpretation of customer information as commercial banks, to maintain a defensive strategy and the specific marketing strategies of the key challenges.In this paper, it is in this context, the use of marketing, management, decision-making theory and methods, data mining and statistical techniques, on a wide range of practical background and prospects for development of the commercial banking customer relationship management core of the issue, to maintain the system study. First in the inspection and analysis on the use of other related variables on the basis that the loss of customers and closely related factors: the length of time in banking services, age, contact with the main channel banks, the Bank of whether to buy certain products, with banks kind of business, such as the number of bank customers and 13 closely related to the loss of the static customer data. In this paper, in a certain bank customers case studies, these static factors, the state inspected at the same time as the introduction of the time-series of factors that inspection period the previous year, the bank transactions, the final part of the two factors as a model, all the input variables remain, the ultimate model of input variables to explain more than 200. Use two data mining software SAS and Weka to set up a commercial bank customers to maintain a decision tree model and forecast Logistic Regression Model, established by the end of the forecast model, to maintain the forecast results are compared and analyzed to identify the commercial banks will be the characteristics of the loss of customers. The results for the commercial banks to maintain bank customers design planning, maintaining valuable customers, improve commercial banks based on the fact that the decision-making ability, with the valuable customer maintain a long-term stable relationship between the increase in profit, the contribution of the banks, and help the banks get real It is the competitive advantage of the theoretical value and practical significance.