PM2.5 concentration forecasting in the area of Jing-Jin-Ji using models based on RF, RR, SVM, and ExtraTrees

Abstract

Abstract In this paper 4 types of machine learning models, i.e., Random forest mode, Ridge regression model, Support vector machine model, and Extremely randomized trees model were adopted to predict PM2.5 based on multi-sources data including air quality, and meteorological data in time series. Data were fed into the model by using rolling prediction method which is proven to improve prediction accuracy in our experiments. The comparative experiments show that at the city level, RF and ExtraTrees models have better predictive results and on season level 4 models all have the best prediction performances in winter time and the worst in the summer time and RF model have the best prediction performance with the IA ranging from 0.93 to 0.98, with an MAE of 5.91 to 11.68 ug/m3. Consequently, the demonstration of models’ different performances in each city and each season is expected to shed a light on environmental policy implications.