Abstract

This study introduces a comprehensive framework for enhancing price forecasting in the oil (Brent and WTI) and natural gas (Henry Hub) markets, which play a critical role in the global economy. By integrating advanced deep learning models and ensemble methods, optimized through Bayesian Hyperparameter Optimization (BO), the research improves predictive accuracy. Utilizing an extensive dataset from January 2010 to February 2024, the models were trained and validated. Results indicate that, in the oil market, the weighted ensemble model combining LSTM and GRU performs best, leveraging the strengths of both models. In the natural gas market, post-optimization, CNN proves most effective in capturing the market's volatility and trends. XGBoost also demonstrates strong performance in both markets, balancing predictive accuracy with training efficiency. These findings offer valuable insights for risk management and decision-making in the energy sector.