Z.V. Maksimenko, D.R. Bogdanova

In the paper a list of the main problems of credit system, including the problems of credit risk management is formed. An approach to the credit risk management, based on a situational management system and client-oriented marketing is suggested. The method of evaluation of credit risk of an individual borrower with the use of artificial neural networks, which allows to find out whether the lending to the borrower is risky or not, is suggested. The method is based on a scoring method of creditworthiness estimation of an individual, which allows fast, objective and impartial estimation of a large amount of potential borrowers and thus reduce the time required for making decision about granting of a loan. It is shown that zero error radial basic network shows the best results of evaluation of credit risk of an individual borrower.