Artificial Neural Networks for Modelling and Predicting Urban Air Pollutants: Case of Lithuania

A study on Vilnius, Lithuania, reveals that sulphur dioxide (SO2) and carbon monoxide (CO) are the primary pollutants affecting air quality in the city's Old Town. Using a 2-layer artificial neural network, researchers achieved high accuracy in predicting and warning about air pollution, identifying SO2 as the most significant pollutant with a normalized importance of 100%, followed by CO at 59%. The study utilized a multilayer perceptron network (MLPN1) with a hyperbolic tangent activation function to analyze data from post-loop traffic implementation in 2020. This approach provides reliable data for air pollution management, addressing the challenges posed by urban expansion and increased industrial and vehicular activity.

Bekesiene, S., & Meidute-Kavaliauskiene, I. (2022). Artificial Neural Networks for Modelling and Predicting Urban Air Pollutants: Case of Lithuania. Sustainability, 14(4), 2470. https://doi.org/10.3390/su14042470