Multivariate Forecasting of Bitcoin Volatility with Gradient Boosting: Deterministic, Probabilistic, and Feature Importance Perspectives
NeutralArtificial Intelligence
- A study has been conducted on the application of the Light Gradient Boosting Machine (LGBM) model for forecasting Bitcoin's realized volatility, utilizing 69 predictors including market and macroeconomic indicators. The research evaluates LGBM's performance against econometric and machine learning models, exploring both deterministic and probabilistic forecasting methods, and highlights key drivers of volatility such as trading volume and investor attention.
- This development is significant as it enhances the understanding of Bitcoin's volatility, providing valuable insights for investors and analysts. The findings may improve forecasting accuracy and inform trading strategies, potentially influencing market behavior and investment decisions.
— via World Pulse Now AI Editorial System
