Advanced Water Quality Assessment through Integrated Machine Learning models for Classification

Water, the basic necessity of any living organisms, constitutes a fundamental human right yet remains tragically inaccessible and unsafe for millions. With nearly 2.5 million annual deaths attributed to waterborne diseases, ensuring access to pollution-free water transcends mere necessity, becoming a potent catalyst for improved global health. Freshwater scarcity is a global problem that is being exacerbated by overexploitation of freshwater sources and climate change impacts. This research delves into machine learning algorithms for analyzing and classifying safe water quality. This research study conducts a comparative analysis of water quality classification models utilizing a series of machine learning algorithms. A series of data preprocessing steps were implemented on this dataset with a view to construct a generic model. This set of steps includes balancing of data using the balancing algorithm and the feature selection. Out of all these applied learning techniques, Random Forest (RF) and XGBoost with hyper parameter (HP) tuning provides the highest accuracy equal to 94.62% and 95.39%.