.png){kind=icon}
.png){kind=icon}
Return
Multivariate bitcoin price volatility forecasting: an optimized decomposition pattern learning model
.png){kind=icon}
PRE
AI
DOI：10.1016/j.irfa.2026.105345.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Developed a multivariate framework for Bitcoin forecasting by integrating EMD/VMD decomposition. • Evolutionary optimization of temporal windows and features improves forecasting and reveals key patterns in volatile markets. • Achieved highly accurate short-term Bitcoin price forecasting using Random Forest (MAPE = 0.01166 ± 0.00105). • Reformulated Bitcoin volatility forecasting as a binary classification problem and achieved 95.47% prediction accuracy. • Demonstrated economic effectiveness and robustness through four-year back-testing under multiple transaction cost scenarios.
Keywords:
Bitcoin
Multivariate
Multistep
EMD/VMD
Volatility patterns
Journal
IF:
9.8
Papers:
4.0K
Citations:
1.9W