Machine Learning algorithms for air pollutants forecasting

Air pollution represents an issue that raises many concerns nowadays, as it has various negative effects on the environment and the economy worldwide. Because of the rapid urbanization, cities are suffering from polluted air, so it is important to predict future air quality. For this purpose, new applications of artificial intelligence should be employed. In this paper, we will present several Machine Learning algorithms, the possible software that can be used for them and the applications used in the field of air quality. Based on the research in the field, we propose SVR, ARIMA and LSTM, 3 Machine Learning models, which can be used to predict air pollution. These algorithms have been tested using time-series for PM
10
and PM
2.5
particles. The results showed that SVR and ARIMA algorithms are the most suitable in forecasting air pollutant concentrations.