Implementation of Machine Learning Models for Predicting the Inactivation Performance of Escherichia Coli in Wastewater Through Varied Photo-Chemical Processes and Aqueous Matrix Combinations

Despite the potential of sunlight-based Photo-chemical Processes (P-CPs) for bacterial inactivation in wastewater, their efficient design requires a delicate balance between optimal bacterial inactivation and the associated costs and time incurred in trial-based designs. To address this challenge, this study explores the implementation of four Machine Learning (ML) models to predict the inactivation performance of E. coli in wastewater across diverse P-CPs. The evaluation of the models demonstrated the usefulness of ML models in streamlining the design process and enhancing the effectiveness of P-CPs without the need for extensive experimental trials. Specifically, the Random Forest model presented the lowest Root Mean Squared Error (RMSE) (0.88) compared to XGBoost, Support Vector Machine (SVM), and Artificial Neural Network (ANN) with values of 1.03, 1.22, and 1.25, respectively.