Research On The Application Of Neural Network Model In Network Credit Risk Assessment

As a complex and nonlinear system,credit risk has some characteristics such as lack of data and asymmetry of distribution,which makes the measurement of credit risk become the difficulty of risk management.Artificial neural network has advantages in credit risk assessment with its strong nonlinear mapping capability and generalization ability,which is suitable for the requirements of credit risk management under the background of Internet finance.P2P lending as a new type of lending model combined with Internet technology and private lending,has made vigorous development in China,but also has been exposed to a lot of problems,such as the recent emergence of P2P lending platform's escape and collapse.An important reason of such events is the platform's capital chain breaks caused by borrower default.Hence,P2P lending platform's development is unsustainable.Credit risk has become the main risk faced by the current P2P lending industry and has threatened the interests of investors as well as the healthy development of P2P lending platform.Therefore,this paper studies the credit risk assessment of P2P borrowers based on the neural network model.Firstly,this paper introduces the basic principle of neural network model and emphatically analyzes the structure and characteristics of BP neural network.It indicates that BP neural network is suitable for borrower credit risk assessment in P2P lending.Secondly,this paper establishes an individual credit risk evaluation index system of P2P lending with reference to the individual credit evaluation system of China's commercial banks.This evaluation system is composed of four dimensions of indicators,including the borrower's personal situation,solvency,repayment willingness and P2P lending platform's certification indicators.After that this paper deals with the data collected from Renrendai platform according to this index system and uses BP neural network to establish the credit risk assessment model for empirical analysis.The training and simulation results of this model show that it can make an accurate prediction of P2P borrower's credit rating.So it can provide reference for the credit risk assessment of P2P network borrowing platform.Finally,this paper puts forward some relevant policy suggestions to improve the credit risk management of P2P lending industry in China according to its current situation.