Pages: 44-51

Abstract

Abstract: Today, one of the main concerns of banks and credit institutions in allocating bank loans and credits is reviewing the risk of credit customers. One of the ways of reducing credit risk and delayed payments can be funding and launching validation companies and institutions. A validation institution is an independent organization which gathers public and legal information, information related to identity, credit transactions and payment records of customers and organizations according to current laws and those that must be implemented. One of the important and main tasks of validation institutions is reviewing the credit information of customers who have visited credit institutions and banks in order to receive loans. This review is in the respect of measuring the rate of customers’ credit and also ranking credit customers. The present research is in the respect of presenting a model to these validation institutions in the respect of ranking regal customers of banks with an approach to multilayer perceptron neural network and logistic regression. In this research, firstly, the effective variables on credit risk have been extracted from credit reports and financial statements and ultimately, the ranking model has been presented through two methods: logistic regression and neural network. And among the two methods of ranking credit customers, the multilayered perceptron artificial neural network method is a more efficient method by considering that it has been more accurate in ranking.