A Hybrid Model for the Prediction of Air Pollutants Concentration, Based on Statistical and Machine Learning Techniques

Abstract

In large cities, the health of the inhabitants and the concentrations of particles smaller than 10 and 2.5 $\upmu$m ($PM_{10}$, $PM_{2.5}$) as well as ozone ($O_3$) are related, making their prediction useful for the government and citizens. Mexico City has an air quality forecast system, which presents a forecast by pollutant at hourly and geographic zone level, but is only valid for the next 24 h.

To generate predictions for a longer time period, sophisticated methods need to be used, but highly automated techniques, such as deep learning, require a large amount of data, which are not available for this problem. Therefore, a set of predictor variables is created to feed and test different Machine Learning (ML) methods, and determine which features of these methods are essential for the prediction of different pollutant concentrations, to develop a hybrid ad-hoc model that includes ML features, but allowing a level of explainability, unlike what would occur with methods such as neural networks.

In this work we present a hybrid prediction model using different statistical methods and ML techniques, which allow estimating the concentration of the three main pollutants in the air of Mexico City two weeks ahead. The results of the different models are presented and compared, with the hybrid model being the one that best predicts the extreme cases.