Forecasting Bank Performance of Private Banks in Iraq Using Neural Networks

Abstract

This paper demonstrate the feasibility of using deep learning approaches in forecasting bank performance using neural networks. The importance of this research is concentrated in the use of an artificial intelligence techniques of neural networks for forecasting bank performance. For that, The performance of (12) banks (research sample) has been evaluated for the period of 16 years using (CAMELS), then forecasting bank performance using neural networks. The research problem is represented in the weak reliance of banks in Iraq on modern technologies for the purpose of predicting their performance. Our results showed that the (NAR) model is capable of predicting bank performance. As well, the (NAR) model is more accurate than time series for the predictions of our performance. Our finding suggest that there is an urgent need for banks to use the (NAR) model to predict their performance due to its predictive ability in Comparison with classic models.