Publication: Malaysian Daily Stock Prediction Analysis Using Supervised Learning Algorithms
Loading...
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
USIM Press
Abstract
Nowadays, Machine Learning (ML) plays a significant role in the economy, especially in the stock trading strategy. However, there is an inadequate extensive data analysis using various ML methods. Previous findings usually focus on the forecasting stock index or selecting a limited number of stocks with restricted features. Therefore, the contribution of this paper focused on evaluating different supervised learning algorithms, namely Logistic Regression (LR), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGB), on a big dataset from 28 stocks in Bursa Malaysia. By setting their parameter along and using Walk-Forward Analysis (WFA) method, the trading signal was evaluated based on Accuracy Rate, Precision Rate, Recall Rate, and F1 Score. For stock trading strategies in Malaysia in particular, the findings of this study show that SVM has a better performance compared to LR and XGB in time series forecasting. The ML algorithms have values ranging from 53% to 66% for Accuracy Rate (AR), Recall Rate (RR), and F1 Score (F1). In addition, SVM has the highest Precision Rate (PR) of 73% among the ML algorithms.
Description
Vol. 8 No. 2 (2022)
Keywords
Machine Learning, Supervised Learning Classifier, Walk-Forward Analysis, time series forecasting
Citation
Halul, H., & Kamil, K. H. (2022). Malaysian Daily Stock Prediction Analysis Using Supervised Learning Algorithms. Malaysian Journal of Science Health & Technology, 8(2), 31–37. https://doi.org/10.33102/2022229