Recurrent neural nets is one of the important deep learning techniques. Efficient Market Hypothesis assumes that financial prices cannot be predicted as they absorb all the news and information. Recently, the enhancement in computational power and learning process by different machine learning techniques facilitates financial market prediction. In this paper, the Long Short Term Memory is followed to predict financial market index prices and returns. The main index of the Egyptian Stock Exchange; EGX30 is investigated. To the best of our knowledge, no research has been conducted to predict the EGX30 using the LSTM model. Two forecasting models were applied; the ARIMA and the Long Short Term Memory for the sake of comparing their performance. The prediction was performed on EGX30 prices and returns. The results show that LSTM performs pretty well in predicting the EGX30 prices and returns. The LSTM performs better than the ARIMA in both prices and returns. Finally, the LSTM model performs better for return than that for prices but overfitting may occur when predicting returns.
Ezzat, H. (2024). Recurrent Neural Networks with LSTM for Stock Market Index Prediction. The International Journal of Informatics, Media and Communication Technology, 6(2), 423-438. doi: 10.21608/ijimct.2024.220153.1048
MLA
Heba M Ezzat. "Recurrent Neural Networks with LSTM for Stock Market Index Prediction", The International Journal of Informatics, Media and Communication Technology, 6, 2, 2024, 423-438. doi: 10.21608/ijimct.2024.220153.1048
HARVARD
Ezzat, H. (2024). 'Recurrent Neural Networks with LSTM for Stock Market Index Prediction', The International Journal of Informatics, Media and Communication Technology, 6(2), pp. 423-438. doi: 10.21608/ijimct.2024.220153.1048
VANCOUVER
Ezzat, H. Recurrent Neural Networks with LSTM for Stock Market Index Prediction. The International Journal of Informatics, Media and Communication Technology, 2024; 6(2): 423-438. doi: 10.21608/ijimct.2024.220153.1048
)
View on SCiNiTO