Data Mining In Ping An Bank Credit Risk Management System

With the development of China's commercial banks and the credit card business, also under the prerequisite of inefficient risk control of clients'credit, the application of data mining technology in the credit card risk management system emerged with great significance. The article elaborates the application of data mining technology in credit risk management system by focusing on the improvement on the key technology of data mining, some functional modules of the data mining system, the integrated application of the credit risk management system, and functional testing and verification for credit risk management system. Meanwhile, the development and prospects of this data mining technology in the industry is also described.In recent years, domestic credit card business has been developing fast, credit card market competition is intensifying. It's very important to establish credit card risk assessment system to increase the market share, fully understanding customer buying behavior and making the correct credit risk strategy. The bank risk audit system which is set up on the data warehouse, online analytical processing, and data mining technology and applies to customer credit card audit system with formation of a group of information application technology, its essence is the intelligence of management activities. Among which, the data warehouse is basic, providing all kinds of information needed for decision making and classification. The decision–maker processes intelligent analysis of customer data warehouse by leveraging online analysis tool. Data mining summarizes the knowledge based on the massive information in data warehouse. Ping An Bank chooses a more popular SLIQ algorithm in the respect of data mining technology, the relative technology improvement on the currently existing SLIQ algorithm, which is the adding of properties priority determination technology and concept, making it better fit in the real business processes and operation, according to real situation during the daily customer audit process, making credit score and customer classification more accurate and efficient, so that algorithms and data mining technology could achieve optimization without compromising the efficiency and accuracy.In the respect of system integration, the paper has a detail description of each function module, meanwhile it describes the structure and key technologies of system integration with a visual map of the system structure. Also the paper makes a detailed description and explanation for the link between the systems in Ping An Bank, the system user interface and business process logic as well.After the verification of the functionality and the feasibility of the key technologies, some functions that the paper is involved are implemented, the paper covers detailed test on multiple functions such as customer automatic approval, automatic refusal, comparison with customer credit record of People's Bank of China and grading on customer credit, it records a detailed verification on the system functionality, after which, also makes a brief description on current system problems and shortcomings , and directions on future improvement.