Penerapan Machine Learning dalam Prediksi Return Saham Syariah di Pasar Modal Indonesia
TL;DR
The findings reveal that the most dominant ML algorithms employed are Support Vector Machines (SVM), Artificial Neural Networks (ANN), and Long Short-Term Memory (LSTM) supported by ensemble methods such as Random Forests and Gradient Boosting, supported by ensemble methods such as Random Forests and Gradient Boosting.
Abstract
This study aims to examine the application of machine learning (ML) in predicting Islamic stock returns in the Indonesian capital market using a Systematic Literature Review (SLR) approach. The literature was collected from Scopus, DOAJ, and Google Scholar databases, covering publications from 2015 to 2025. The findings reveal that the most dominant ML algorithms employed are Support Vector Machines (SVM), Artificial Neural Networks (ANN), and Long Short-Term Memory (LSTM), supported by ensemble methods such as Random Forests and Gradient Boosting. Recent research trends also indicate a shift toward hybrid approaches, integrating ML algorithms with conventional technical indicators to enhance prediction accuracy while maintaining relevance to established market analysis practices. However, a significant gap exists between research practices in Indonesia and global trends. Research in Indonesia remains limited to index forecasting and portfolio optimization, whereas globally ML has been applied more broadly in areas such as Islamic credit scoring, anomaly detection and anti-money laundering (AML), Shariah-compliant robo-advisors, and data-driven zakat and waqf management. This gap is primarily driven by limited data/AI literacy, minimal integration of non-traditional data, and the absence of AI governance frameworks that incorporate Shariah compliance principles. Therefore, future research is recommended to focus on developing AI governance frameworks based on maqashid al-shariah, integrating alternative data such as NLP-based sentiment analysis and blockchain, and diversifying ML applications in the Indonesian Islamic capital market into areas such as halal credit scoring, Shariah compliance automation, and data-driven ZISWAF management. These findings are expected to contribute not only to improving prediction accuracy but also to strengthening investor trust and enhancing the global competitiveness of Indonesia’s Islamic capital market.