A Literature Survey: AQI Prediction Using ML, DL and Hybrid Models

Abstract The existence of human being is not worth considering without fresh air. Rapidurbanization and industrialization is degrading air quality, which puts people’shealth at risk by causing various lung and heart diseases. It is necessary to raisepublic awareness to safeguard their well-being from these detrimental effects.An effective solution that must be implemented in conjunction with awarenesscampaigns are the use of modelling and monitoring systems for air quality predictions.Many air quality monitoring stations are installed to monitor and analysethe air quality of the respective regions. Many Machine Learning (ML) andDeep Learning (DL) models have shown their effectiveness to enhance predictivecapabilities and address the challenges of air pollution forecasting. This articleprovides a comprehensive survey based on thorough analysis of several ML,DL and hybrid models presented by the various researchers for air quality prediction.The selected papers are categorized based on the models used by theresearchers. This comprehensive survey of air quality prediction is an asset thatprovides insights that can greatly help to address the pressing problems broughton by poor air quality and its monitoring. In this survey diverse pollutants, meteorologicalfactors, several ML models, DL models and hybrid models with theirperformance efficacy are being analysed thoroughly to report the effect of thesemodels in the area of air quality forecasting. This survey facilitates the creationof realistic and efficient plans to protect public health from the adverse effects ofair pollution.