Water quality in Delhi is a pressing environmental and health issue, with significant variation in water quality across different city regions. This study aims to predict and analyze the ranking Water quality using machine learning models based on pollutants such as pH, TSS (Total Suspended Solids), COD (Chemical Oxygen Demand), BOD (Biochemical Oxygen Demand), Oil and Grease, Ammoniacal Nitrogen as N, Sulphide, Phosphate, Total Kjeldahl Nitrogen (TKN). Data were collected from government monitoring stations across twelve key regions: Lawrence Road, Mangolpuri, Wazirpur, Badli, Nangloi, Naraina, Okhla, Narela, Mayapuri, GTK, Jhilmil, and SMA. Four classification algorithms-Support Vector Machine (SVM), XGBoost, Random Forest, and Logistic Regression-were applied to assess their effectiveness in ranking the area according to water quality. After thorough preprocessing and feature selection, model performance was evaluated using rank co-relation. XGBoost achieved the highest accuracy (98.6%), followed by Random Forest (88.2%), showing strong capability in modeling complex patterns in water quality ranking data. SVM and Logistic Regression had lower predictive performance. Region-wise analysis ranked Lawrence road as having the worst water quality, followed by nangloi, while SMA recorded the best water quality, with mangolpuri in a moderate range. These results demonstrate the potential of ensemble learning models in predicting and mapping urban air quality, providing actionable insights for policymakers to implement targeted environmental and public intervention.