A Novel Interpretable Machine Learning Model Approach for the Prediction of TiO 2 Photocatalytic Degradation of Air Contaminants

Abstract

Abstract Air contaminants lead to various environmental and health issues. Titanium dioxide (TiO2) features the benefits of autogenous photocatalytic degradation of air contaminants. To evaluate its performance, laboratory experiments are commonly used to determine the kinetics of the photocatalytic-degradation rate, which is labor intensive, time-consuming, and costly. In this study, Machine Learning (ML) models were developed to predict the photo-degradation rate constants of air-borne organic contaminants with TiO2 nanoparticles and ultraviolet irradiation. The hyperparameters of the ML models were optimized, which included Artificial Neural Network (ANN) with Bayesian optimization, gradient booster regressor (GBR) with Bayesian optimization, Extreme Gradient Boosting (XGBoost) with optimization using Hyperopt, and Catboost combined with Adaboost. The organic contaminant was encoded through Molecular fingerprints (MF). A data collection protocol was developed to collect data from published papers to analyze the effects of a variety of contributing factors on the photo-catalytic degradation performance. A large portion of data was found missing values. Imputation methods were used to estimate the missing values that allow the dataset to be fully utilized. Furthermore, a generative ML model Vanilla Gan model was utilized to create synthetic data to further augment the size of available dataset. The results indicated that data imputation allowed to fully utilize the limited amount of available dataset to achieve good ML prediction performance and prevent problems such as overfitting common with small-sized data. Besides, augmentation of experimental data with synthetic data significantly improved the ML prediction accuracy and reduced considerably the overfitting problems. The SHapley Additive exPlanations (SHAP) was employed for ML model interpretability. The results ranked the feature importance and assessed the impacts of different experimental variables on the rate of photo-degradation, which were consistent with physico-chemical laws. Interpretable ML models allows to shed light on the mechanism and present a promising tool in the prediction and assessment of the major contributing factors on the TiO2 photo-degradation rate of air contaminants. Besides direct findings on the potentials of ensembled ML models for TiO2 photocatalytic performance prediction, this study showed that imputation processes in data pre-processing to fill missing values and generative ML model for data augmentation allow to fully utilize the value of data, which is important for successful application of ML model for small and imperfect dataset commonly seen in engineering and science domains.