Air Quality Monitoring and Pollution Prediction Using Machine Learning

Air pollution poses a significant threat to public health, ecosystems, and urban sustainability. Accurate monitoring and prediction of air quality are essential for proactive environmental management. This study presents a machine learning-based approach for predicting the Air Quality Index (AQI) using pollutant and meteorological data collected from the Okhla monitoring station in Delhi, India. The dataset spans three years and includes 16 features such as PM2.5, PM10, NO ${ }_{2}, \text{CO}$, Benzene, temperature, humidity, and wind speed. We applied and compared seven supervised machine learning algorithms: Random Forest, Decision Tree, k-Nearest Neighbors (k-NN), Support Vector Machines (SVM), Gradient Boosting (XGBoost), Naive Bayes, and Artificial Neural Networks (ANNs). Advanced feature selection methods including Lasso, CFS, and RFE were used to enhance model performance. Evaluation metrics such as accuracy, precision, recall, and $\mathbf{F 1}$-score were calculated, with XGBoost and Random Forest achieving the highest overall performance ($\mathbf{F 1}$-score above $\mathbf{0. 8 3}$). The proposed models demonstrate strong potential for real-time AQI forecasting and can support timely public health advisories and policy decisions.