The Mining Of Characteristics Of The Internet Financial Marketing Customer Accurate Classification

Since the reform and opening up, Chinese residents’ disposable income continued to increase, but the return on investment for Chinese residents as well as the utilization of idle funds investments in developed countries are relatively low, residents of personal financial needs and urgent.The birth of the Internet financial break this limit, as long tail crowd solve financial difficulties to provide new ideas and programs. In the current rapid development of Internet financial situation, which involved all of its customers are in a constant process of exploration and adaptation, and is gradually formed a relatively stable behavior and manner. Identifying their answers to these questions to understand customer behavior, rational allocation of client resources, so that precise marketing, improve their products, improve customer service quality, improve business performance, enhance the market competitiveness of enterprises of great significance.In this study, the Internet banking platform for customer classification feature selection perspective, a financial circles in Hong Kong stocks FORMAX traffic data for the data source, the customer classification feature that the impact of important customer information indicator marketing end result excavation. In the method, this study selected the most representative of the current statistical learning model in five categories, including logistic regression(LOGISTIC), K nearest neighbor(KNN), linear discriminant analysis(LDA), quadratic discrimi--nant analysis(QDA), random forest(RANDOMFOREST). We use FORMAX financial circles in Hong Kong stocks data related to customer information, combined with the above statistical classification model, preferably a kind of optimal model, empirical analysis and comparative analysis, the corresponding analytical results, and financial circles in Hong Kong stocks FORMAX proposed business outlook.After comparative analysis model used in this study as the Random Forest method selection model, and its model optimization. The present study suggests that random forest model for the current financial circles in Hong Kong stocks FORMAX business marketing success discriminant analysis fit well and can continue to optimize the future can be determined in advance and do precision marketing at the influx of a large number of customers.