Machine Learning Models for Drinking Water Quality Classification

Access to safe drinking water is essential for living. It is important to keep track of its components through periodical water quality monitoring. Water quality can be defined as the collective measure of its chemical, biological, and physical components. Water contamination can come from human activities such as the discharge of untreated waste from factories, the use of harmful pesticides, or natural causes such as earthquakes. In the recent past, water quality classification has gained the attention of researchers in Machine Learning. This has been attributed to the advancement of computing resources and the availability of big data. In this paper, Machine Learning algorithms were trained using the Water Quality dataset from the Indian Government website through Kaggle. Water Quality Index was used to classify the quality of the data. Decision Trees, Multilayer Perceptron (MLP), XGBoost, K-Nearest Neighbor, and Support Vector Machine were the ML algorithms studied. The Precision, Accuracy, Recall, and F1-Score were utilized as metrics to evaluate the performance of the models. With an accuracy of 95.12%, the results demonstrated that XGBoost was a superior classifier for water quality, followed closely by Support Vector Machine with an accuracy of 93.22%