What Is the Effect of Urban Traffic and Meteorology on Air Pollution? Norwegian Evidence Using Machine Learning Approaches

This paper uses machine learning approaches to explore the association between traffic volume, air pollution, and meteorological conditions. A focus is the interaction between these factors. The paper does this using hourly data for Oslo, Norway. I considered a total of ten datasets to verify the prediction performance of the model. I found that ARMAX outperformed support vector machines and decision trees in air pollution prediction. At the same time, I also explored the effect of dividing the seasons and weather subsets on forecast accuracy. Finally, my study makes optimal policy recommendations for reducing air pollution and further health-related burdens from traffic, after considering the interaction and lag effects of meteorology, time variables, traffic, and air pollution