| With the development of the 'Internet Plus',we have met the big data revolution.Internet finance has broken its traditional monopolized pattern?P2P network platform?the third-party payment and crowdfunding have sprung up.But our countries'credit system still has room for improvement,especially in personal consuming credit.Though the P2P network platform has strong demand for personal credit,not only can they connect to the People's Bank of China,but also did they get information of the borrower from credit information center,which greatly increased the operational risk.This paper aims to reduce operational risk and provide some advice for the P2P platform through the borrower' credit research.First,this paper discussed the meaning of risk management in theoretical aspect and explained the economical meaning for building personal credit model.Meanwhile,It analysed the feasibility of K-Means clustering model and Bayesian discriminant model.Second,under the background of?Interim Measures for the Administration of the Business Activities of Online Lending Information Intermediary Institutions?,this paper analysed the current situation and future development trend of P2P platform.besides,this paper demonstrates the normal risk management pattern in our country.Then,this paper did empirical research on the borrower' credit risk management.I collected 6471 actual borrower' information data in Renrendai.At first,I made clustering analysis and divided the credit grade into five categories and made the loan decision standard;then this paper made feature selection and found the borrowers'age,income,house and working time are more influencial to clustering,So I build the Bayesian discriminant model using this index and found its accuracy is 87.4%,which is an ideal result.Finally,this paper discriminated the new sample of Renrendai platform?At last,based on the actural operation and the prediction model,this paper give some guidance and suggestions to the P2P network platform. |
PDF Full Text Request