The Application Of Decision Tree Algorithm In Credit Risk Assessment Of P2P Net Loan

P2P online lending originated in Britain, with the gradual development of Internet banking, has become a new model of finance, it has played an important role in the individual loans and financing of small and medium-sized enterprises. However,with the dramatically growth, the online lending industry will also faced with significant credit risk, and credit risk assessment is the decisive factor to effectively identify and reduce the credit risk. In its starting step,P2 P credit model faces many advantages as well as varies challenges in China.,especially the enormous credit risk..Based on the P2 P online credit transaction data, this paper uses decision tree to build a credit risk assessment model to analyze. In this experiment, according to a known sample of the original credit rating status, based on historical data for training and testing,it can identify several indicators from a range of possible indicators and default individuals(including natural persons, legal persons and other organizations).Then, in turn, based on these warning in advance, it can identify and predict the likelihood of future violations of other individuals(including natural persons, legal persons and other organizations), so as to take appropriate measures to mitigate risks in advance.The main work of this paper includes:(1) In order to research the background and significance, we first discuss the domestic and foreign development status of this project, including the academic level and existing problems,in order to generate the outline and content of this paper;(2) It describes the basic theory of credit risk assessment. First defines four basic concepts of credit risk assessment.which contains credit risk, credit risk management and credit risk assessment.In order to clarify theoretical system of credit risk assessment theory;(3) Based on the researches of the credit risk system, it constructs the evaluation of indicator system of P2 P online lending credit risk,.The performance and results of the model has been analyzed in this paper. Eventually, by using actual data on credit risk assessment indicator system and algorithm based on C4.5,it makes credit risk assessment model to be improved and Empirical.