A review on emerging artificial intelligence (AI) techniques for air pollution forecasting: Fundamentals, application and performance

Accurate air quality forecasting is critical for systematic pollution control as well as public health and wellness. Most of the traditional forecasting techniques have shown inconsistent predictive accuracy due to the non-linear, dynamic and complex nature of air pollutants. In the past few years, artificial intelligence (AI)-based methods have become the most powerful and forward-looking approaches for air pollution forecasting because of their specific features such as organic learning, high precision, superior generalization, strong fault tolerance, and ease of working with high-dimensional data. This study presents a comprehensive overview of the most widely used AI-based techniques for air pollution forecasting namely Artificial Neural Networks (ANN), Deep Neural Network (DNN), Support vector machine (SVM) and Fuzzy logic through a systematic literature review (SLR). In total 90 papers were selected which were distributed between 2003 and 2021. The SLR aims to classify the literature on AI-based air pollution forecasting from various perspectives, such as input parameters, relative frequency of application of AI techniques, performance, year of publication, journal and geographic distribution and also addresses the corresponding research questions related to this domain. The results showed that the number of citations and publications have been increasing in recent years. The most frequently applied input parameter is the air quality and the best performing AI-based technique is the DNN. On the other hand, Fuzzy logic, DNN and SVM are the three commonly used AI-based techniques for air pollution forecasting. In addition, some technological gaps in the literature and the pros and cons associated with the different AI techniques, were identified and discussed. This review article shows that AI-based techniques have triggered a resurgence of interest in air pollution forecasting and offer great potential to fundamentally change the way air pollution is forecasted in the near future.