A comparative study on multi-regression analysis and BP neural network of PM2.5 index

The air pollution PM2.5 index is affected by a variety of factors. The research shows that the PM 2.5 index of yesterday, precipitation of yesterday, average temperature, average humidity, precipitation and wind velocity are the main factors which can affect PM2.5 index. The paper utilizes the data of surface meteorological observation and air pollution PM2.5 index of Wuhan City from November 1, 2013 to January 31, 2014. We construct the air pollution PM2.5 index forecasting model according to multiple regression analysis and BP neural network. The empirical study shows that compared to regression prediction model, BP neural network prediction model obtains the nonlinear relation among the prediction factors after training. The predictive model based on BP neural network is of higher predictive precision and better prediction effect which can be used in kinds of prediction research and has good value to popularize.