Comparison of Performance Classical and Neural Networks Models for Estimation Credit risk and Capacity Customers: Evidence from Tejarat bank

As an important goal، financial institutions in order to enhance their performance identify customers to credit allocation to those who are less likely to default. But for this purpose some common methods such as personal judge، analysis and audit have been used. However، most of these methods have focused on credit risk of customers، while the credit capacity to provide facilities for customers can play an important role to implement. Therefore، this paper uses neural network model to calculate both the credit risk factor and capacity at the same time. Simultaneous، linear and logistic regression models to calculate the credit risk and capacity has been compared with the neural networks model results. Results imply higher efficiency of neural networks than linear regression to estimate the capacity and efficiency of credit customers.