Research On The Credit Assessment Of Peer To Peer Lending Borrower Based On Fuzzy Clustering

Peer to peer lending is a financial model in which both the supplier and the demander of funds?investors and borrowers?use the online lending platform as the intermediary for unsecured lending.It is the product of the integration of network technology and private lending.P2P lending has the characteristics of convenience,speed and low threshold,and plays an important role in alleviating the difficulty of small and micro customers'borrowing and promoting the development of inclusive finance.In recent years,the domestic P2P lending business has developed rapidly,but at the same time,the network lending platform generally has the problems of borrower's credit risk is too large,and the ability of credit risk control is weak.A large number of borrowers of online loans default or overdue behavior,resulting in many online lending platforms shut down or bankruptcy.How to accurately evaluate the credit level of borrowers and improve the credit risk management ability of online lending platform is an urgent problem to be solved.Firstly,the paper introduces the concepts of credit risk and credit evaluation of borrowers in online lending platform,and analyses the applicability of fuzzy clustering and fuzzy pattern recognition methods in the field of credit evaluation of borrowers in online lending platform and credit grade identification of new borrowers.Secondly,on the basis of the available indicators of Renrendai website,and drawing lessons from the credit evaluation indicators and basic framework of mainstream credit evaluation institutions at home and abroad,this paper establishes a credit evaluation index system for borrowers of online lending platform with 16 indicators from 7 indicators,including borrower's basic information X₁,borrowing situation X2,working condition X3,personal assets X4,credit record X5,certification information X₆ and macroeconomic situation X₇.Then,based on the fuzzy clustering method and the entropy weight method,this paper constructs the credit evaluation model of the borrowers of the online lending platform,and establishes the credit grade recognition model of the new borrowers of the online lending platform by using the approach degree method of the fuzzy pattern recognition.Finally,an empirical study is carried out with1000 loan records of Renrendai website.The results show that:firstly,the importance of each credit evaluation index layer is obtained by using the method of entropy weight,that is,working condition X₃>borrowing condition X₁>macroeconomic condition X₆>credit record X7>asset information X4>basic information X2>authentication information X5.Secondly,in terms of credit evaluation model,F test is used to verify the significance of classification.The results show that the model has good classification significance.Thirdly,in the aspect of credit grade recognition of new borrowers of online credit,the credit grade subordination of new borrowers of online credit is determined according to the approach degree method of fuzzy pattern recognition,thus realizing the recognition and prediction of credit grade of new borrowers of P2P online credit platform.This paper focuses on the credit evaluation of borrowers in the P2P lending model,constructs the credit evaluation model of borrowers in the P2P lending platform and the credit identification model of new borrowers in the online lending platform,and tests the effect of the model by empirical analysis.Finally,some suggestions are put forward,such as establishing a comprehensive and scientific credit evaluation index system and strengthening the application of big data technology,which can be used as a reference for decision-making of P2P lending platforms and relevant institutions.