AquaAugmentor Boosts Water Potability Prediction Accuracy

Access to safe drinking water is a global imperative, yet accurately classifying water quality remains challenging due to the inherent complexity and variability of water source data. This research addresses the problem of predicting water potability using machine learning and deep learning algorithms, particularly for low-dimensional datasets where feature scarcity can limit model performance. The core contribution is "AquaAugmentor," a novel feature augmentation algorithm specifically designed to enhance the predictive capabilities of these models. The study leverages a dataset comprising various chemical attributes of water, such as pH, hardness, solids, chloramines, and sulfate. It systematically evaluates the performance of different machine learning and deep learning models both with and without the application of AquaAugmentor. The models classify water as either potable or non-potable, and their performance is rigorously compared using metrics like test accuracy and AUC score. The findings consistently highlight that AquaAugmentor significantly improves predictive performance, offering valuable insights into effective techniques for enhancing water quality classification. This work aims to support researchers, policymakers, and public health officials in making more informed decisions regarding safe water access and environmental quality assessments.