Water-Quality Prediction Based on H2O AutoML and Explainable AI Techniques

Rapid expansion of the world’s population has negatively impacted environment, notably water quality. As a result, water-quality prediction arisen as hot issue during last decade. Existing techniques fall short in terms good accuracy. Furthermore, presently, dataset available for analysis contains missing values; these values have significant effect on performance classifiers. An automated system that deals with efficiently and achieves accuracy is proposed this study. To handle problem, study makes use stacked ensemble H2O AutoML model; to values, KNN imputer. Moreover, compared seven machine learning algorithms. Experiments are performed two scenarios: removing using The contribution each feature regarding explained SHAP (SHapley Additive exPlanations). Results reveal model outperforms other models 97% accuracy, 96% precision, 99% recall, 98% F1-score prediction.