Predicting water potability using machine learning classification methods

Abstract

Ensuring access to safe drinking water remains a critical global challenge, making the ability to assess water potability both efficiently and accurately increasingly important. This study explores the use of supervised machine learning techniques to predict the potability of water based on a set of physicochemical attributes, including pH, hardness, turbidity, dissolved solids, and various chemical concentrations. Two classification models are examined: logistic regression, which provides a linear and interpretable decision boundary, and Support Vector Machines (SVM), which offer greater flexibility by capturing complex, non-linear relationships within the data. A systematic evaluation framework is employed to compare both models using key performance indicators such as accuracy, recall, precision, and F1-score. These metrics allow for a comprehensive understanding of each model’s strengths, limitations, and capacity to generalize to unseen samples. The analysis aims not only to identify the more effective algorithm but also to highlight the potential of machine learning as a reliable tool for environmental monitoring and water quality assessment. The findings contribute to ongoing efforts to automate and enhance potability prediction, thereby supporting informed decision-making in public health and resource management.