A Hybrid Deep Ensemble Learning Framework for Cryptocurrency Price Prediction

Mohammadreza  Haghighi; Seyed Yashar Banihashem; Mohammadmehdi  Gilaniansadeghi

Authors

Mohammadreza Haghighi Department of Information and Technology Management, Qa.C., Islamic Azad University, Qazvin, Iran Author https://orcid.org/0000-0001-9428-9842
Seyed Yashar Banihashem Assistant Professor, Department of Electrical and Computer Engineering, Buein Zahra Technical University, Buein Zahra, Iran https://orcid.org/0000-0002-2456-7920
Mohammadmehdi Gilaniansadeghi Department of computer engineering, Qa.C., Islamic Azad University, Qazvin, Iran Author https://orcid.org/0009-0002-5108-6433

Keywords:

Cryptocurrency Price Prediction, Deep Ensemble Learning, CNN, LSTM, GRU, Transformer, Stacking, Time, Series Forecasting, Machine Learning

Abstract

This study aimed to develop and evaluate a hybrid deep ensemble learning framework integrating CNN-LSTM, GRU, and Transformer architectures through stacking for accurate next-day cryptocurrency price prediction. This quantitative predictive study analyzed daily market data for five major cryptocurrencies, including Bitcoin, Ethereum, Binance Coin, Ripple, and Cardano, over the period from January 1, 2018, to December 31, 2025. The dataset comprised 14,610 cryptocurrency-day observations, with 2,922 observations for each asset. Data were chronologically divided into training, validation, and testing subsets. Historical price, trading volume, return, volatility, lagged variables, and technical indicators were used as predictors. A 30-day sliding window was employed to predict the following day’s closing price. The proposed framework combined CNN-LSTM, GRU, and Transformer models using a stacking-based meta-learner. Performance was evaluated using Mean Absolute Error, Root Mean Squared Error, Mean Absolute Percentage Error, and R². Diebold-Mariano tests were used to compare out-of-sample forecast accuracy. The proposed hybrid ensemble achieved an MAE of 0.0294, RMSE of 0.0436, MAPE of 3.12%, and R² of 0.971, outperforming CNN-LSTM, GRU, Transformer, ARIMA, Random Forest, XGBoost, and persistence benchmarks. Diebold-Mariano tests showed that the ensemble produced significantly lower forecast loss than all competing models, including Transformer, the strongest standalone model, with all p values below 0.001. Cryptocurrency-specific analyses yielded R² values above 0.960 for all five assets, while volatility-regime analysis showed that predictive accuracy declined under high-volatility conditions but remained strong, with an R² of 0.952. The findings demonstrate that combining convolutional, recurrent, and attention-based deep-learning mechanisms through stacking can significantly improve the accuracy and robustness of cryptocurrency price forecasting across heterogeneous assets and changing volatility conditions.

Downloads

Download data is not yet available.

References

1. Sattarov O, Choi J. From Prediction to Profit: A Comprehensive Review of Cryptocurrency Trading Strategies and Price-Forecasting Techniques. IEEE Access. 2024;12:87039-63. doi: 10.1109/ACCESS.2024.3417449.

2. Qureshi SM, Saeed A, Ahmad F, Khattak AR, Almotiri SH, Al Ghamdi MA, et al. Evaluating machine learning models for predictive accuracy in cryptocurrency price forecasting. PeerJ Computer Science. 2025;11:e2626. doi: 10.7717/peerj-cs.2626.

3. Derbentsev V, Datsenko N, Stepanenko O, Bezkorovainyi V, editors. Forecasting Cryptocurrency Prices Time Series Using Machine Learning. M3E2-MLPEED; 2019.

4. Chowdhury R, Rahman A, Rahman S, Mahdy MRC. An approach to predict and forecast the price of constituents and index of cryptocurrency using machine learning. Physica A: Statistical Mechanics and its Applications. 2020;551:124569. doi: 10.1016/j.physa.2020.124569.

5. Parvini N, Fazilzadeh A, Jabbarzadeh Y. Cryptocurrency Price Prediction Using AI -Based Forecasting Models: University of Tabriz; 2019.

6. Varastehpour S, Bakhit S, Salem M, Rabiei M, Abdol Abadi A, editors. Forecasting the Price of the First Virtual Cryptocurrency Based on Machine Learning Techniques2019: Shahid Beheshti University.

7. Bakhit S, Rabiei M. Forecasting the Price of the First Virtual Cryptocurrency Based on Deep Machine Learning Techniques: Ghiyasuddin Jamshid Kashani University; 2020.

8. Khonsarian F, Timourpour B, Rostegar MA. Forecasting Price and Portfolio Selection Model for Financial Assets and Cryptocurrencies: Tarbiat Modares University; 2020.

9. Kumar R, Singh P. Optimizing LSTM Networks for Cryptocurrency Price Forecasting. Information Processing & Management. 2020;57(6):102337. doi: 10.1016/j.ipm.2020.102337.

10. Anderson R, Patel M. Cryptocurrency Price Prediction Using CNN and LSTM NetworksJO - International Journal of Forecasting. 2020;36(4):1476-90.

11. Gonzalez M, Sanchez F. Time Series Forecasting of Cryptocurrency Prices Using CNNs and RNNs. Journal of Computational Finance. 2019;22(3):120-36.

12. Ma J, Zhou W. Hybrid CNN-RNN Models for Forecasting Cryptocurrency Prices. Neurocomputing. 2019;342:66-73.

13. Shivaei V, Pazhouhan MR, Yazdian Dehkordi M. Forecasting Cryptocurrency Price Using Hybrid Machine Learning Methods: Yazd University; 2019.

14. Xiaolei S, Mingxi L, Zeqian S. A novel cryptocurrency price trend forecasting model based on LightGBM. Finance Research Letters. 2020;32:101084. doi: 10.1016/j.frl.2018.12.032.

15. Bouteska A, Abedin MZ, Hajek P, Yuan K. Cryptocurrency price forecasting - A comparative analysis of ensemble learning and deep learning methods. International Review of Financial Analysis. 2024;92:103055. doi: 10.1016/j.irfa.2023.103055.

16. Alnami H, Mohzary M, Assiri B, Zangoti H. An Integrated Framework for Cryptocurrency Price Forecasting and Anomaly Detection Using Machine Learning. Applied Sciences. 2025;15(4):1864. doi: https://doi.org/10.3390/app15041864.