Review of neural network models for air quality prediction

Abstract

With the diversified development of air quality data acquisition and processing technology, people began to gradually adopt the new technology represented by machine learning to predict the air quality data to make up for the shortcomings of traditional forecasting methods. However, many machine learning models applied to air quality prediction generally use batch learning and prediction methods, that is, after a sample study and prediction, new samples will not be learned, and air quality prediction will be increased. The error, which deviates from the track of real-time prediction, is difficult to apply effectively to actual engineering. Therefore, in view of the problems existing in the current air quality prediction field, we review the previous research on the air quality neural network prediction model.