Certain Investigation of Machine Learning Algorithms in Prediction of Drinking Water Quality

Water is essential for all living organisms, which covers almost seventy percent of the surface of Earth. Nowadays water is polluted mainly due to human intervention activities such as industrial waste disposal, agricultural runoff, urban development, and improper sewage management significantly contribute to the degradation of water quality and the disruption of aquatic ecosystems. Implementing a Reverse Osmosis (RO) mechanism in institutions is essential for reducing pollution, particularly from hazardous gases. This study provides a classification model to predict water quality utilizing learning algorithms like Bagging, AdaBoost, Decision Tree, and K-Nearest Neighbor (KNN). The dataset contains potability, sulfate, chloramines, conductivity, solids, organic carbon, trihalomethane, turbidity, hardness, and pH values. By comparing all the algorithms, the proposed methodology obtained a better prediction with the help of the randomized searchCV optimization technique. All these algorithms have been considered highly accurate with an accuracy of 67%, 66%, 67.9%, and 66% respectively. It aims to forecast the quality of water to ensure a sustainable future.