Artificial intelligence-enhanced optimization for methyl paraben removal from aqueous solutions using green-synthesized copper oxide nanoparticles on magnetic polyglucosamine/alginate biocomposites

This study examines the removal of methyl paraben (MP) from aqueous solutions using green-synthesized copper oxide nanoparticles supported on magnetic chitosan/alginate biocomposites (CuO-MCAB). The upgraded functional conditions were recognized as pH = 9.00 (X1), MP concentration = 1 mg/L (X2), contact time = 90 min (X3), and adsorbent dosage = 1 g/L (X4), accomplishing a removal efficiency of 91 %. Different Artificial Intelligence (AI) models were assessed to predict and advance model performance. Among the models, Extreme Gradient Boosting (XGBoost) showed predominant performance with R 2 = 0.95 for the train data and R 2 = 0.84 for the test data. Also, the values of the minimal errors of this model were presented such as Root Mean Square Error (RMSE) = 0.09, Mean Absolute Error (MAE) = 0.08, and Mean Squared Error (MSE) = 0.01 on the test dataset. Markedly, other models like Support Vector Regression (SVR) had worse precision with the SVR having a R 2 0.92 (train) and 0.78 (test), RMSE and MAE of 0.15 and 0.12, respectively. Feature importance analysis using SHapley Additive exPlanations (SHAP) and feature importance plots uncovered that X4 had a strong effect on removal efficiency, while the interaction between X1 and X3 likewise also played an important role in improving removal performance. The present research shows how effective this combination is to include the green synthesis methods with the advanced AI model that offers an efficient and ecologically friendly method to clear the natural impurities of the water reservoirs. The findings give a good basis on creating the imaginative innovations in unpacking the water.