沙漠与绿洲气象2025，Vol.19Issue(3)：193-200,8.DOI:10.12057/j.issn.1002-0799.2304.25001
基于聚类分型的随机森林O3浓度预测方法研究
Research on O3 Concentration Prediction Using Random Forest Cluster Classification
摘要
Abstract
This study proposed a random forest prediction model optimized with a fuzzy C-means clustering algorithm.The model utilized monitoring data for six air pollutants(O3,PM2.5,PM10,NO2,SO2,CO)along with weather forecast data from 2014 to 2020.Initially,two clustering factors were identified through cross-correlation analysis.O3 concentrations were then classified into three categories using the fuzzy C-means clustering algorithm.A random forest model was subsequently constructed to predict O3 concentrations,with its performance evaluated both before and after clustering.The results indicate that the previous day's O3 and PM10 concentrations have the most significant impact on the next day's O3 levels,and seasonal variations also play a critical role.Following fuzzy C-means clustering,the Mean Absolute Error(MAE)and Root Mean Square Error(RMSE)of the predicted O3_8 h concentrations decreased by 10.5%and 8.8%,respectively.Additionally,the coefficient of determination(R2)increased,confirming an improvement in prediction accuracy.These findings highlight the practical value of the proposed model for forecasting O3 pollution in Shanghai.关键词
O3/空气污染物/模糊C均值/聚类分型/随机森林/机器学习Key words
O3/air pollutant/fuzzy C-means/cluster classification/random forest/machine learning分类
天文与地球科学引用本文复制引用
韩晶晶,迪里努尔·牙生,雷雨虹,尚子溦,田瑜,王金艳..基于聚类分型的随机森林O3浓度预测方法研究[J].沙漠与绿洲气象,2025,19(3):193-200,8.基金项目
上海市浦东新区民生科研专项(PKJ2021-N04) （PKJ2021-N04）
国家重点研发计划项目(2020YFA0608402) （2020YFA0608402）
甘肃省自然科学基金(21JR7RA501,21JR7RA497) （21JR7RA501,21JR7RA497）
