Perbandingan Performa Model Prediksi Volatilitas BTC/IDR Menggunakan LSTM dan ARIMA
DOI:
https://doi.org/10.30865/json.v7i4.9719Keywords:
ARIMA, BTC/IDR, LSTM, Volatilitas, Time SeriesAbstract
Karakteristik fluktuatif pasar aset kripto yang ekstrem menuntut ketersediaan model peramalan yang andal sebagai penunjang strategi manajemen risiko investasi. Penelitian ini bertujuan untuk membandingkan pendekatan Long Short-Term Memory (LSTM) sebagai model deep learning sekuensial dan Autoregressive Integrated Moving Average (ARIMA) sebagai model statistik deret waktu dalam memprediksi log-volatility Bitcoin pada pasangan BTC/IDR periode 2018–2025. Dataset historis harian BTC/IDR diperoleh dari platform Binance dengan periode observasi Januari 2018 hingga Desember 2025, kemudian diproses melalui perhitungan log-return, estimasi realized volatility berbasis jendela 7 hari, transformasi logaritmik, serta normalisasi data. Evaluasi model menggunakan metode walk-forward validation dengan metrik Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), dan koefisien determinasi (R²). Hasil penelitian menunjukkan bahwa model LSTM memperoleh MAE sebesar 0,5126, RMSE sebesar 1,0408, dan R² sebesar 0,6803, sedangkan model ARIMA menghasilkan MAE sebesar 0,5430, RMSE sebesar 1,0217, dan R² sebesar 0,7052 pada konfigurasi terbaiknya. Meskipun LSTM memiliki MAE yang lebih rendah, model ARIMA menunjukkan performa yang lebih unggul berdasarkan nilai RMSE yang lebih kecil dan R² yang lebih tinggi, sehingga lebih efektif dalam menjelaskan variasi data serta menangkap fluktuasi ekstrem pada volatilitas Bitcoin. Secara keseluruhan, hasil penelitian menunjukkan bahwa model ARIMA lebih representatif dalam memodelkan dinamika log-volatility Bitcoin dibandingkan model LSTM. Temuan ini menegaskan bahwa pemilihan model prediksi volatilitas perlu mempertimbangkan karakteristik data yang dinamis dan fluktuatif. Penelitian ini diharapkan dapat memberikan kontribusi dalam pengembangan model prediksi volatilitas yang adaptif, khususnya pada pasar cryptocurrency di Indonesia.
References
[1] Y. Liu et al., “Risks and Returns of Cryptocurrency,” 2018. [Online]. Available: https://ssrn.com/abstract=3226952
[2] B. Dharma, P. Gusniati, and T. Wardani, “ANALISIS PEMANFAATAN CRYPTOCURRENCY BITCOIN SEBAGAI ALAT ALTERNATIF INVESTASI,” 2023.
[3] M. Zahid, F. Iqbal, and D. Koutmos, “Forecasting Bitcoin Volatility Using Hybrid GARCH Models with Machine Learning,” 2022, doi: 10.3390/risks.
[4] E. Bouri, A. A. Salisu, and R. Gupta, “The predictive power of Bitcoin prices for the realized volatility of US stock sector returns,” Financial Innovation, vol. 9, no. 1, Dec. 2023, doi: 10.1186/s40854-023-00464-8.
[5] G. Dudek, P. Fiszeder, P. Kobus, and W. Orzeszko, “Forecasting cryptocurrencies volatility using statistical and machine learning methods: A comparative study,” Appl. Soft Comput., vol. 151, Jan. 2024, doi: 10.1016/j.asoc.2023.111132.
[6] Z. C. Huang, I. Sangiorgi, and A. Urquhart, “Forecasting Bitcoin volatility using machine learning techniques,” Journal of International Financial Markets, Institutions and Money, vol. 97, Dec. 2024, doi: 10.1016/j.intfin.2024.102064.
[7] Y. Wang, G. Andreeva, and B. Martin-Barragan, “Machine learning approaches to forecasting cryptocurrency volatility: Considering internal and external determinants,” International Review of Financial Analysis, vol. 90, Nov. 2023, doi: 10.1016/j.irfa.2023.102914.
[8] N. A. Sofiah, K. D. Tania, A. Meiriza, and A. Wedhasmara, “A Comparative Assessment SARIMA and LSTM Models for the Gurugram Air Quality Index’s Knowledge Discovery,” in 2024 International Conference on Electrical Engineering and Computer Science (ICECOS), 2024, pp. 26–31. doi: 10.1109/ICECOS63900.2024.10791243.
[9] F. A. Miranda, K. D. Tania, and R. D. Kurnia, “Comparative Performance Evaluation of ARIMA, SARIMA, and LSTM for Daily Shallot Price Forecasting in Palembang City,” Electronic Journal of Education, Social Economics and Technology, vol. 6, no. 2, p. 1323, Dec. 2025, doi: 10.33122/ejeset.v6i2.1323.
[10] M. L. Pratama and H. Utama, “PENDEKATAN DEEP LEARNING MENGGUNAKAN METODE LSTM UNTUK PREDIKSI HARGA BITCOIN,” IJCSR: The Indonesian Journal of Computer Science Research, vol. 2, no. 2, 2023, doi: 10.37905.
[11] N. Urooj, L. Asif, and Z. Jabin, “Bitcoin Price Forecasting: A Comparative Study of Machine Learning, Statistical and Deep Learning Models,” 2024.
[12] A. Kareem and A. Aue, “Bitcoin Forecasting with Classical Time Series Models on Prices and Volatility,” Nov. 2025, [Online]. Available: http://arxiv.org/abs/2511.06224
[13] C. Alexander, D. F. Heck, and A. Kaeck, “The Role of Binance in Bitcoin Volatility Transmission,” Appl. Math. Finance, vol. 29, no. 1, pp. 1–32, 2022, doi: 10.1080/1350486X.2022.2125885.
[14] P. Anand and A. M. Sharan, “Bitcoin Return Dynamics Volatility and Time Series Forecasting,” International Journal of Financial Studies, vol. 13, no. 2, Jun. 2025, doi: 10.3390/ijfs13020108.
[15] S. H. Sung, J. M. Kim, B. K. Park, and S. Kim, “A Study on Cryptocurrency Log-Return Price Prediction Using Multivariate Time-Series Model,” Axioms, vol. 11, no. 9, Sep. 2022, doi: 10.3390/axioms11090448.
[16] J. Chevallier and B. Sanhaji, “Jump-Robust Realized-GARCH-MIDAS-X Estimators for Bitcoin and Ethereum Volatility Indices,” Stats (Basel)., vol. 6, no. 4, pp. 1339–1370, Dec. 2023, doi: 10.3390/stats6040082.
[17] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735.
[18] S. Kim, “Innovating knowledge and information for a firm-level automobile demand forecast system: A machine learning perspective,” Journal of Innovation and Knowledge, vol. 8, no. 2, Apr. 2023, doi: 10.1016/j.jik.2023.100355.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Jurnal Sistem Komputer dan Informatika (JSON)

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

This work is licensed under a Creative Commons Attribution 4.0 International License
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under Creative Commons Attribution 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (Refer to The Effect of Open Access).

