A survey of data-driven insights into air quality trends and prediction in India: Proceedings of the 4th International Conference on Computational Methods in Science & Technology (ICCMST 2024), 2–3 May 2024, Mohali, India, Volume 1

Air pollution is a major health and environmental issue in increasingly urbanizing countries like India. This study examines ML and DL implementations on Indian Air Quality Index (AQI) datasets from 2015 to 2020. The research seeks to identify patterns, anticipate trends, and guide air pollution mitigation decisions. The study carefully analyses AQI parameters, explores data, and uses powerful ML and DL algorithms to forecast AQI and identify pollution hotspots. Technical features include missing value imputation, feature engineering, model selection criteria, hyper parameter tuning procedures, and model performance assessment measures. Key results show that ML and DL models can accurately forecast air quality, identify pollution sources, and capture temporal and geographical fluctuations. The article suggests integrating more data sources, exploring advanced deep learning methods, and developing dynamic air quality control systems. This study uses data-driven insights and multidisciplinary techniques to improve our knowledge of air quality dynamics in India and create the framework for proactive air pollution mitigation and public health policies.