Air Quality Prediction and Influencing Factor Identification Based on Machine Learning Methods
The accurate prediction of air quality index(AQI)and the identification of influencing factors are of great practical significance for air pollution prevention and control.The AQI of Beijing from the first quarter of 2014 to the second quarter of 2022 was selected as the research object to explore the influence of six major pollutants,five meteorological factors and fourteen economic variables on air quality.The DT,RF,GBDT and XGBoost models were selected to predict AQI,and the contribution of each variable to AQI was quantitatively analyzed using the stability selection method.The results show that the four model methods have excellent prediction effects,and XGBoost and RF have the best prediction effects;among the six major pollutants,PM2.5,PM10 concentration and meteorological factors,such as wind speed and pressure,have a greater influence on AQI;the influence of fourteen economic variables on AQI is quite different,among which the per capita disposable income of urban residents,tertiary industry GDP and gross industrial output value above designated size have a greater influence on AQI,while the primary industry GDP and road cargo transportation volume have a small influence.
air qualityinfluencing factorsquantitative analysismachine learningselection of stability
国家科技期刊平台
NSTL
万方数据
维普
