Comparison of Forecasting the Index Price Movement in Financial Institutions using Artificial Intelligence

This study predicts the movements in the stock price index of Tehran Stock Exchange by using neural networks. The source of this paper is the information from banks and financial institutions listed on the Tehran Stock Exchange during the years 1385 to 1391 are used. The results show that the ANFIS algorithm has the best performance between FA, RBF, MLP and ICA algorithms. Results indicate that the proposed algorithms overall have high ability to predict the stock price index movement of Tehran Stock Exchange. Output of MATLAB shows that correlation coefficient of ANFIS algorithm about 0.9985.