Enhancing Water Quality Prediction with Machine Learning and Explainable AI Techniques

Different pollutants have affected the quality of water in the recent past in a big way. Clean drinking water is important for every living creature on this planet because the second important natural resource, water, can be used in many ways for instance in agriculture, to water livestock, or drink. Proposed methodology ensures water quality by applying various computational algorithms like Decision Trees, K-Nearest Neighbor, RandomForest, XGBoost, CatBoost, Logistic Regression, stacking classifier. Synthetic Minority Over-sampling Technique has used to increase the size of minority class and avoid an over fitted model Light-gbm has given the best accuracy of 96 %. The different methods used in this work include Explainable AI (XAI) methods like SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) are used to make the results understandable to user. This helps in identifying the key features responsible for determining the water quality such as pollutant levels, mineral content. The proposed model provides a user friendly graphical user interface which allows the user to access and interpret the results and indicates weather the water is safe for consumption or not.