Publication

Abstract : Exposure to air pollution can significantly affect one’s health, which can lead to itchy eyes, breathing issues, and hospitalisation. There are a variety of pollutants such as indoor, biological, and outdoor pollutants. This paper emphasizes estimating the quality of air concerning outdoor pollutants. This model aims to effectively predict the Air Quality Index (AQI) in major populated stations in Bengaluru city. Using data from five monitoring stations - Peenya, Bapuji Nagar, Silk Board, Hombegowda, and BTM Layout- collected between 2021 and 2023, machine learning models are developed to forecast AQI for 2024. The research involved data pre-processing, AQI calculation from key pollutants, and training of models for AQI prediction. Our findings indicate that the Random Forest algorithm outperformed the Decision Tree in both regression and classification tasks, providing more accurate predictions. Additionally, the study included a visual mapping of AQI categories using colour codes, revealing trends in air quality across different regions. This research serves as a foundation for future work, including expanding the model to predict AQI across India and integrating it with real-time data collection hardware to enhance prediction accuracy.