Machine Learning Techniques for Predicting the Egyptian Stock Exchange

Document Type : Original Article

Author

Department of Socio-Computing, Faculty of Economics and Political Science, Cairo University, Cairo, Egypt

Abstract

Predicting future movement of stock prices is very significant for researchers, investors, and companies. Due to the sensitivity of financial markets to any news, the market absorbs updates instantly. This hardens the forecasting process pretty much. However, efforts exerted in this field reveal that machine learning methods are very promising in predicting stock time series data. Many advances are achieved by implementing different machine learning techniques. Machine learning methods have improved ability of learning and training from datasets to produce more accurate forecasting. In this paper, three different machine learning methods are implemented; Prophet, K-Nearest Neighbor, and Feedforward Neural Network. EGX30, the main index of the Egyptian Stock Exchange, was collected for about 25 years to be modelled by the proposed three methods. To the best of our knowledge, no research has been conducted to predict the Egyptian Stock Exchange with these three models. Accuracy metric was reported to compare the performance of the three models. Surprisingly, Prophet model performed the best though it is not so famous in predicting stock prices.

Keywords

Main Subjects

Files

Share

How to cite

Statistics