1.png)
Peningkatan Prediksi Harga Saham Menggunakan Correlation-Driven PCA dan Temporal Attention-Based Deep Learning
Abstract
Fluktuasi harga saham yang tinggi menjadikan prediksi pergerakan harga saham sebagai tantangan yang kompleks. Berbagai pendekatan berbasis indikator teknikal telah digunakan, namun sering kali kurang efektif dalam menangkap pola nonlinier dan korelasi antar variabel. Penelitian ini mengusulkan integrasi Correlation-Driven Principal Component Analysis (CD-PCA) dengan Temporal Attention-Based Deep Learning sebagai metode untuk meningkatkan kualitas prediksi harga saham. CD-PCA digunakan untuk mentransformasi 1.195 indikator teknikal menjadi komponen utama yang mempertimbangkan korelasi terhadap variabel target, sehingga mampu mengurangi redundansi antar fitur tanpa kehilangan informasi penting. Data yang digunakan adalah saham perbankan besar Indonesia, yaitu BBCA, BBNI, BBRI, dan BMRI, dengan periode perdagangan dari 1 Januari 2015 hingga 30 Desember 2024. Eksperimen dilakukan dengan tiga skenario pembagian data, yaitu 70:30, 80:20, dan 90:10. Hasil penelitian menunjukkan bahwa model yang diusulkan mampu mengurangi hasil error prediksi secara konsisten. Pada skenario 70:30, saham BBCA mencatatkan MAPE terendah sebesar 2,17% dengan R² mencapai 0,85, menunjukkan performa prediksi yang baik ketika data latih cukup besar. Sementara itu, BBNI dan BBRI justru memberikan performa terbaik pada skenario 90:10 dengan MAPE masing-masing 3,56% dan 3,41%, serta R² sebesar 0,85 dan 0,87. Sebaliknya, BMRI menunjukkan tingkat error lebih tinggi dan R² negatif pada skenario 70:30, menandakan kompleksitas pola harga saham yang sulit ditangkap. Secara keseluruhan, penelitian ini menegaskan bahwa integrasi CD-PCA dan Temporal Attention-Based Deep Learning berkontribusi terhadap pengurangan error prediksi harga saham, khususnya pada saham BBCA, BBNI, dan BBRI, sehingga berpotensi menjadi kerangka kerja andal dalam peramalan pasar saham di Indonesia.
Keywords
Full Text:
PDFReferences
Al-Khasawneh, M. A., Raza, A., Khan, S. U. R., & Khan, Z. (2024). Stock Market Trend Prediction Using Deep Learning Approach. Computational Economics, 0123456789. https://doi.org/10.1007/s10614-024-10714-1
Billah, M. M., Sultana, A., Bhuiyan, F., & Kaosar, M. G. (2024). Stock price prediction: comparison of different moving average techniques using deep learning model. Neural Computing and Applications, 36(11), 5861–5871. https://doi.org/10.1007/s00521-023-09369-0
Botunac, I., Bosna, J., & Matetić, M. (2024). Optimization of Traditional Stock Market Strategies Using the LSTM Hybrid Approach. Information, 15(3), 136. https://doi.org/10.3390/info15030136
Chen, C.-H., Chen, P.-Y., & Chun-Wei Lin, J. (2022). An Ensemble Classifier for Stock Trend Prediction Using Sentence-Level Chinese News Sentiment and Technical Indicators. International Journal of Interactive Multimedia and Artificial Intelligence, 7(3), 53. https://doi.org/10.9781/ijimai.2022.02.004
Haryono, A. T., Sarno, R., & Sungkono, K. R. (2024). Stock price forecasting in Indonesia stock exchange using deep learning: a comparative study. International Journal of Electrical and Computer Engineering (IJECE), 14(1), 861. https://doi.org/10.11591/ijece.v14i1.pp861-869
He, X., & Wang, J. (2024). A Hybrid Forecasting System Based on Comprehensive Feature Selection and Intelligent Optimization for Stock Price Index Forecasting. Mathematics, 12(23), 3778. https://doi.org/10.3390/math12233778
Ji, S. (2023). Application research of SGD algorithm based on PCA dimensionality reduction technique for stock price prediction. 2023 International Conference on Electronics and Devices, Computational Science (ICEDCS), 265–269. https://doi.org/10.1109/ICEDCS60513.2023.00054
Khanpara, P., Kadam, R., Lavingia, K., & Patel, S. (2023). Stock Trend Prediction: A Comparative Study using Different Approaches. Proceedings - 5th International Conference on Smart Systems and Inventive Technology, ICSSIT 2023, Icssit, 1697–1701. https://doi.org/10.1109/ICSSIT55814.2023.10060936
Li, Z., Lang, L., Sun, G., Cai, Z., & Luo, Z. (2023). Enhancing Multiple Linear Regression for Price Prediction: A PCA-Integrated Approach. 2023 4th International Conference on Computer, Big Data and Artificial Intelligence (ICCBD+AI), 337–341. https://doi.org/10.1109/ICCBD-AI62252.2023.00063
Mansoor, Z. K., & Al-Sultan, A. Y. (2025). Enhancing Stock Price Prediction Using Mutual Information, PCA, and LSTM: A Deep Learning Approach. Fusion: Practice and Applications, 17(1), 196–208. https://doi.org/10.54216/FPA.170114
Mi, Y., Xu, D., & Gao, T. (2023). Application of PCA-LSTM algorithm for financial market stock return prediction and optimization model. International Journal for Simulation and Multidisciplinary Design Optimization, 14(8), 1–6. https://doi.org/10.1051/smdo/2023009
Ming, L., & Chen, G. (2024). Stock Price Prediction Based on Relative Strength Index, Moving Average Convergence Divergence and XGBoost Model. 2024 IEEE 13th Data Driven Control and Learning Systems Conference (DDCLS), 1988–1993. https://doi.org/10.1109/DDCLS61622.2024.10606580
Priyatno, A. M., Ningsih, L., & Noor, M. (2024). Harnessing Machine Learning for Stock Price Prediction with Random Forest and Simple Moving Average Techniques. Journal of Engineering and Science Application, 1(1), 1–8. https://doi.org/10.69693/jesa.v1i1.1
Priyatno, A. M., Ramadhan Sudirman, W. F., & Musridho, R. J. (2024). Feature selection using non-parametric correlations and important features on recursive feature elimination for stock price prediction. International Journal of Electrical and Computer Engineering (IJECE), 14(2), 1906. https://doi.org/10.11591/ijece.v14i2.pp1906-1915
Priyatno, A. M., Sudirman, W. F., Musridho, R. J., & Amalia, F. (2023). Impurity-Based Important Features for feature selection in Recursive Feature Elimination for Stock Price Forecasting. Jurnal Teknik Industri Terintegrasi, 6(4), 1182–1194. https://doi.org/10.31004/jutin.v6i4.17726
S, J., & Arya, K. (2024). Stock Movement of Wipro Using Technical Analysis. 2024 10th International Conference on Advanced Computing and Communication Systems (ICACCS), 1, 1033–1038. https://doi.org/10.1109/ICACCS60874.2024.10717324
Sheng, Y., Fu, K., & Wang, L. (2022). A PCA-LSTM Model for Stock Index Forecasting: A Case Study in Shanghai Composite Index. 2022 7th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA), 412–417. https://doi.org/10.1109/ICCCBDA55098.2022.9778913
Uçkan, T. (2024). Integrating PCA with deep learning models for stock market Forecasting: An analysis of Turkish stocks markets. Journal of King Saud University - Computer and Information Sciences, 36(8), 102162. https://doi.org/10.1016/j.jksuci.2024.102162
Uma, K. S., & Srinath Naidu. (2021). Prediction of Intraday Trend Reversal in Stock Market Index Through Machine Learning Algorithms. In Advances in Intelligent Systems and Computing: Vol. 1200 AISC (pp. 331–341). https://doi.org/10.1007/978-3-030-51859-2_30
Wang, J., Liu, D., Jin, L., Sun, Q., & Xue, Z. (2023). A PCA-IGRU Model for Stock Price Prediction. Journal of Internet Technology, 24(3), 621–629. https://doi.org/10.53106/160792642023052403007
Zhu, C., Lu, P., Feng, W., & Wang, Y. (2025). Bimodal Stock Price Prediction Based on Holt-Winters Exponential Smoothing and PCA Whitening Transformation. IAENG International Journal of Computer Science, 52(1), 187–200.
DOI: https://doi.org/10.17509/ijdb.v5i3.90659
Refbacks
- There are currently no refbacks.
Copyright (c) 2025 Universitas Pendidikan Indonesia (UPI)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Indonesian Journal of Digital Business is published by Universitas Pendidikan Indonesia (UPI)
and managed by Department of Digital Business
Jl. Dr. Setiabudi No.229, Kota Bandung, Indonesia - 40154
View My Stats