Designing Efficient and Sustainable Predictions of Water Quality Indexes at the Regional Scale Using Machine Learning Algorithms

Water quality and scarcity are key topics considered by the Sustainable Development Goals (SDGs), institutions, policymakers stakeholders to guarantee human safety, but also vital protect natural ecosystems. However, conventional approaches deciding suitability of water for drinking purposes often costly because multiple characteristics required, notably in low-income countries. As a result, building right trustworthy models is mandatory correctly manage available groundwater resources. In this research, we propose check classification techniques such as Decision Trees (DT), K-Nearest Neighbors (KNN), Discriminants Analysis (DA), Support Vector Machine (SVM), Ensemble (ET) design best strategy allowing forecast Quality Index (WQI). To achieve goal, an extended dataset characterized samples collected total twelve municipalities Wilaya Naâma Algeria was considered. Among them, 151 were examined training samples, 18 used test confirm prediction model. Later, data classified based on WQI into four states: excellent quality, good poor very or unsafe water. The main results revealed that SVM classifier obtained highest accuracy, with 95.4% accuracy when standardized 88.9% samples. confirmed use machine learning powerful tools forecasting larger scales promote efficient sustainable control support decision-plans.