AIR QUALITY PREDICTION USING MACHINE LEARNING ALGORITHMS FOR ENVIRONMENTAL MONITORING AND PUBLIC SAFETY

Abstract

Air quality has become a major concern due to rapid urbanization, industrialization, and increasing vehicular emissions, affecting human health and the environment. This paper presents an air quality prediction system using a Hybrid Air Quality Prediction Algorithm (HAQP) that combines multiple machine learning techniques. The system utilize environmental parameters such as PM2.5, PM10, NO₂, SO₂, CO, O₃, temperature, and humidity. Data preprocessing methods including missing value handling, noise removal, and feature selection are applied to improve performance. The hybrid model integrates algorithms such as Linear Regression, Random Forest, Support Vector Machine (SVM), Decision Tree, and Gradient Boosting. The model is evaluated using MAE, RMSE, and R²score. Experimental results show that the system provides accurate predictions and supports real-time environmental monitoring and public safety.