PUT ARTICLE TITLE HERE (22)

This study compares some of the previously described methods, which combine the empirical mode decomposition (EMD) and penalized quantile regression.These methods are EMD-RQR, EMD-LQR, and EMD-EnetQR.These methods provide more accurate estimations to tackle multicollinearity, heavy-tailed distributions, heterogeneity, and selecting the most significant variables.Simulation studies and real example were applied.The results showed that the EMD-LQR method outperformed other methods by effectively identifying the decomposition components, with high prediction accuracy, followed by the EMD-EnetQR method.The findings suggest that integrating EMD with these regression techniques enhances model robustness and predictive performance.Moreover, the EMD-LQR and EMD-EnetQR effectively identify significant decomposition components with the most significant impact on the response variable, enhancing the analysis of relationships between EMD decomposition components and the response variable.While the EMD-RQR method selected all decomposition components in the final model.