Electricity price forecasting with ensemble meta-models and SHAP explainers: a PCA-driven approach

Key Points

• Electricity price forecasting improved accuracy and reliability, demonstrating significant benefits in prediction outcomes.
• The primary metric of interest is forecasting accuracy, with a noted reduction of prediction errors under this model.
• Analysis using PCA-driven ensemble meta-models provides detailed insights into the data contributing to predictions for better understanding.
• The effectiveness of SHAP explainers indicates potential for enhanced transparency in predictive analytics, especially in financial sectors.