This analysis demonstrates improved forecasting accuracy in bitcoin prices using wavelet transform and neural networks, suggesting potential for better models.
Bitcoin price forecasting is challenging due to its volatility and complex patterns. This study proposes a Wavelet Transformed DBN-BiGRU Hybrid Model, combining Discrete Wavelet Transform (DWT), Deep Belief Networks (DBNs), and Bidirectional Gated Recurrent Units (Bi-GRUs) to enhance predictive accuracy. The model achieved MAE = 1.447, RMSE = 0.524, MAPE = 1.092%, and R² = 0.992, outperforming traditional models. DWT isolates noise, DBNs extract features, and Bi-GRUs model temporal dependencies, enabling superior forecasting. While effective, limitations include sensitivity to wavelet selection and computational cost. Future work should explore adaptive wavelets and external data integration for further improvements.
No takes yet. Share an insight, caveat, or question.
Jiawei Zhou (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: