Application Of Decision Tree Algorithm In Breach Prediction

In 1997, The People’s Bank of China began to build credit registration and consulting system for banks(formerly corporate credit information system). At the end of November 2013, the credit system included more than 830 million individual records, businesse and other organization records nearly 20 million. So far, China has credit report based credit evaluation system have a solid foundation. The research and analysis for the credit report is not much, I believe this situation will change with opening the personal credit market In the second half. Credit market developments abroad is relatively mature earlier, many well-known rating agency decision tree algorithm as the core of the applicant’s credit status assessment, decision tree algorithm developed earlier, the algorithm is stable, and the resulting decision rules easier to understand, to analyze many variable type, for large data sets. For all of these advantages, this paper hopes to decision tree algorithm as the core, hopes to analyze the data in credit reports listed in the field, to identify the user a high default risk, we have reached the warning lending risk.