Water quality detection and analysis using machine learning algorithms to assess the impact of influencing factors on health issues

Water quality classification is essential due to increased contamination risks from industrial growth and chemical discharge. This study presents a supervised machine learning classification framework for assessing drinking water safety and identifying major influencing factors. Using a Kaggle dataset comprising nine key water attributes, we implemented and evaluated classification algorithms including Light Gradient Boosting Machine (LGBM), Decision Tree, Histogram Gradient Boost Machine (HGBM), and Random Forest. The Random Forest classifier achieved the highest accuracy of 83.33%, substantially outperforming previous systems on the same data, which reported accuracy ranges between 60% and 70%. The proposed system, implemented in Python with a Flask-based interface, provides users with a direct classification of water samples as safe or unsafe and indicates the primary factor responsible when water is unsafe. These results demonstrate that supervised classification methods can reliably support water quality assessment and aid public health protection.