A Novel Hybrid Model to Forecast Crude Oil Price Based on Hybrid Kernel Extreme Learning Machine and Multi-Objective Chaos Game Optimization

As a key industrial raw material, fluctuations in crude oil prices have a profound impact on the global economy. Accurately forecasting crude oil prices is critical for ensuring national energy security and maintaining stable economic operations. However, the high complexity, volatility, and nonlinear characteristics of crude oil prices pose significant challenges to accurate forecasting. To address these challenges, this work proposes an innovative hybrid forecasting framework that integrates signal decomposition, a hybrid kernel extreme learning machine, chaos theory, and multi-objective swarm intelligent optimization. The model decomposes and reconstructs the original crude oil data to effectively mitigate the adverse effects of negative noise on forecasting. It then develops a hybrid kernel extreme learning machine model based on the multi-objective chaos game optimization algorithm to effectively capture the global features of crude oil price volatility, while simultaneously improving both forecasting accuracy and robustness. To evaluate the model’s effectiveness, this study uses WTI and Brent crude oil price datasets for empirical analysis. The experimental results demonstrate that the proposed hybrid model exhibits superior performance in both prediction accuracy and stability, confirming its effectiveness and practical utility as a tool for crude oil price forecasting.