Application of Artificial Intelligent in the Prediction of Credit Rating of Banks Customers

This study examines a multi-objective fuzzy simplex-genetic algorithm being developed to predict the financial performance of legal customers of banks. Predicting the performance produced by the model was examined based on its ability to accurately identify credit default. Using available data from Keshavarzi bank for the period between 2001-2006, debt ratio, operational ratio, and return on equity were selected as descriptive variables, and on the other side, dependent variable was considered as a dummy variable. In order to train and validate the model, data were divided into two sets, model (in-sample) and test (out-of-sample). After running the algorithm, regardless the sensitivity and specificity ratios, the key variable were specified.