MLP Enhanced CO Emission Prediction Model with LWSSA Nature Inspired Optimization.

Environmental degradation due to the rapid increase in CO₂ emissions is a pressing global challenge, necessitating innovative solutions for accurate prediction and policy development. Machine learning (ML) techniques offer a robust approach to modeling complex relationships between various factors influencing emissions. Furthermore, ML models can learn and interpret the significance of each factor's contribution to the rise of CO. This study proposes a novel hybrid framework combining a Multi-Layer Perceptron (MLP) with an enhanced Locally Weighted Salp Swarm Algorithm (LWSSA) to address the limitations of traditional optimization techniques, such as premature convergence and stagnation in locally optimal solutions. The LWSSA improves the standard Salp Swarm Algorithm (SSA) by incorporating a Locally Weighted Mechanism (LWM) and a Mutation Mechanism (MM) for greater exploration and exploitation. The LWSSA-MLP framework achieved a prediction accuracy of 97% and outperformed traditional optimizer-based MLP models across several evaluation metrics. A permutation feature significance analysis identified global trade, coal energy, export levels, urbanization, and natural resources as the most influential factors in CO₂ emissions, offering valuable insights for targeted interventions. The study provides a reliable and scalable framework for CO₂ emission prediction, contributing to actionable strategies for sustainable development and environmental resilience.