<![CDATA[Crude oil and coal are fundamental energy sources for global industrialization. The stability of their prices plays an important role in supporting global economic growth. The increasing demand for crude oil and coal to meet the energy needs of various countries significantly affects price fluctuations in the global market. Crude oil and coal price volatility also exhibits a strong positive relationship, where shocks in the oil market propagate to the coal market. Therefore, reliable and accurate predictive models are essential for forecasting the price movements of these energy commodities. This study compares the performance of two Recurrent Neural Network (RNN) models, namely Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), an approach not previously undertaken. Predictions were performed on daily time-series data covering the period from 2015 to 2026. Model performance was evaluated using Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Root Mean Square Error (RMSE) and . The optimal parameters used in this study were 64 units, a dropout rate of 0.2, batch size of 32, 100 epochs, and a 60-step time window. The results indicate that the GRU model outperformed the LSTM model in terms of prediction accuracy and computational time. The proposed predictive model can support energy market risk analysis, investment planning, and policy evaluation related to the stability of energy commodity prices.]]>
Copyrights © 2026
