|
|
 |
| |
|
|
|
Using Hybrid Models for Predicting Volatility in the Saudi Stock Market |
|
|
|
PP: 349-363 |
|
|
doi:10.18576/jsap/150304
|
|
|
|
Author(s) |
|
|
|
Najla F. Al-Qadi,
Salwa S. B. Saad,
Osama M. H. Mahdi,
Fatima Eljaili,
Asia Bashier,
Rafia Zaroug,
Abdelgalal O. I. Abaker,
|
|
|
|
Abstract |
|
|
| The ability to predict volatility in the Saudi Stock Exchange Index (TASI) is of great importance in supporting investment decisions and risk management. Understanding daily volatility behavior contributes to enhancing market efficiency and improving trading strategies. This study aims to build models capable of classifying high and low volatility based on the variables of opening price and trading volume, using a combination of statistical models and artificial intelligence algorithms. Four main models were applied: logistic regression, multi-layer neural network (MLP), hybrid model (soft voting), and decision tree. The results showed that the neural network model outperformed in prediction accuracy and high recall and F1 values. The hybrid model is expected to achieve a good balance between different statistical metrics, thanks to its combination of the simplicity of linear models and the ability of neural networks to capture non-linear patterns. The study confirms that the use of hybrid models represents a promising direction in improving the accuracy of volatility prediction in financial markets |
|
|
|
|
|
|
|
