Research on Prediction of Air Quality Index Based on NARX and SVM

Abstract

In view of the problem of increasingly serious air pollution nowadays, this paper adopts Nonlinear Auto Regressive models with Exogenous Inputs (NARX) and Support Vector Machine (SVM) for regression and prediction on the Air Quality Index (AQI). The results show that the mean squared error of train set, validation set and test set and correlation coefficient of train set, validation set and test set in NARX model are 0.0029, 0.0064, 0.0072, 0.9865, 0.9731, 0.9701, respectively. Meanwhile, the results in SVM model are 0.0119, 0.0347, 0.0159, 0.9083, 0.9612, 0.8891, respectively. The above data prove the advantage of the NARX method in the prediction of the AQI.