Tue, Aug 5, 2025
[Archive]
Institute of Physics and Technical science, L.N. Gumilyov Eurasian National University, Astana, 010000, Kazakhstan
Abstract: (50 Views)
This study aims to provide valuable insights into future emission trends by utilizing advanced predictive modeling techniques. With global energy consumption continuing to rise, understanding and forecasting carbon dioxide (CO2) emissions from energy sources is crucial for policymakers to design effective mitigation measures and transition towards sustainable energy systems. Predicting energy-related CO2 emissions is vital for informing evidence-based environmental policies and strategies to combat climate change. This project investigates the prediction of energy-related carbon dioxide emissions in Western Europe by merging a neural network with three nature-inspired optimization algorithms: Multiverse Optimization (MVO), League Championship Algorithm (LCA), and Evaporation Rate Water Cycle Algorithm (ERWCA). We assess how much this combined approach improves prediction accuracy using a relevant dataset. Our findings demonstrate that the ensemble model works better than alternative methods and has increased accuracy in estimating carbon dioxide emissions, as evaluated by R-squared (R2) and Root Mean Square Error (RMSE). This research provides helpful information for developing sustainability initiatives and regulations by highlighting the advantages of utilizing various optimization techniques in predictive modeling for environmental applications. The accuracy of the MLP is improved by applying the MVO, LCA, and ERWCA algorithms. It was demonstrated that some hybrid techniques can yield more precise predictions than those derived from the conventional MLP ranking. Subsequent analysis revealed that ERWCA outperforms the other algorithms. Using R2 = 0.9977 and 0.9919, RMSE 17.9936 and 30.1394 for ERWCA, R2 = 0.9962 and 0.9898, RMSE 23.3505 and 33.8724 for MVO, and R2 = 0.9898 and 0.9793, RMSE 38.2511 and 48.1272 for LCA, the CO2 emission was estimated with the highest degree of accuracy.
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
