Accurate prediction of atmospheric pollution is critical for public health, guiding environmental policies, and mitigating the adverse effects of air pollution. Traditional statistical models and standalone machine learning algorithms, while useful, often fail to capture the complex, nonlinear interactions between multiple factors influencing air quality, such as meteorological conditions, traffic emissions, and industrial activities. This paper presents a comprehensive review of machine learning techniques applied to air pollution prediction, with a special focus on the growing trend of hybrid models (HM). In addition, this paper highlights future research directions centered on developing adaptive HM capable of integrating diverse data streams, addressing gaps in data availability, and dynamically responding to changing pollution patterns. Furthermore, the paper presents a strategy on how combining machine learning algorithms can enhance predictive accuracy and robustness by leveraging the unique capabilities of each model. The findings from this study aim to provide a foundation for future research and practical applications in air quality management, ultimately contributing to more effective pollution forecasting and control strategies.
Zouglis et al. (Mon,) studied this question.