Air Quality Index Prediction Based on Multilayer Perceptron

Abstract

Taking the comprehensive evaluation index AQI as the research goal, the data of Beijing is studied, and an AQI prediction model based on multilayer perceptron neural network (MLP neural network) is proposed. First, the Beijing air quality dataset is constructed, and the AQIs from 2014 to 2019 are extracted as input features. Then, the air quality index of Beijing in 2020 is predicted based on MLP. After that, through parameter adjustment and model comparison, the final prediction accuracy reached 0.861.