Air quality prediction algorithm based on ARIMA-SVM combined model

Abstract

The invention relates to an air quality prediction algorithm based on an ARIMA-SVM combined model, and belongs to the field of data mining. The algorithm comprises the following steps: firstly, checking the stability of air quality data; if the data is not stable, allowing the data to be stable through differential processing; and performing ARIMA modeling analysis according to the characteristicsof the data yt to obtain a prediction result, performing prediction analysis on the residual Nt of the prediction result by using an SVM model to obtain a result, and adding the prediction results ofthe two models to obtain a prediction result of the ARIMA-SVM combined model. According to the combined model provided by the invention, the problem that the prediction precision is reduced due to the fact that linear and nonlinear feature information of data cannot be mined at the same time when air quality data is analyzed and predicted by an existing single model can be improved, and the prediction precision is greatly improved.

Description

technical field [0001] The invention relates to the field of data mining, in particular to an air quality prediction algorithm based on an ARIMA-SVM combined model. Background technique [0002] In recent years, with the development of industrial production and the increase of human activities, a large amount of energy consumption and waste discharge have been caused, and the problem of air quality has become increasingly prominent, especially the inhalable particulate matter (PM2.5), which seriously affects human health. Therefore, the accurate prediction of air quality has important guiding significance for people's production, life, and cultivation of environmental awareness. [0003] The concentration of PM2.5 is affected by many factors and has both linear and nonlinear characteristics. At present, the prediction of PM2.5 mainly includes neural network, gray prediction, spatio-temporal data model, support vector machine and other methods. The above existing single mode...

AI Technical Summary

Problems solved by technology

The above existing single models cannot simultaneously mine the linear and nonlinear feature information of the data when analyzing and predicting air quality data, resulting in a decline in prediction accuracy.

Images