基于K-均值聚类和贝叶斯判别的城市空气质量等级分类及预测  

Classification and Prediction of Urban Air Quality Grade Based on K-means Clustering and Bayesian Discrimination

作　　者：常丽娜[1] 王颖俐[1] 王瑶 CHANG Lina;WANG Yingli;WANG Yao(The Department of Mathematics,Changzhi University,Changzhi 046011,China)

机构地区：[1]长治学院数学系,山西长治046011

出　　处：《太原师范学院学报（自然科学版）》2021年第2期41-46,共6页Journal of Taiyuan Normal University:Natural Science Edition

基　　金：山西省高等学校教学改革创新重点项目(J2020320);国家级大学生创新项目(2019588);长治学院校级改革创新项目(JC201910).

摘　　要：为了研究城市的空气质量情况,对65个重点环保城市2018年的样本数据进行分析,在利用K-均值聚类将样本重点环保城市划分等级的基础上,进一步利用贝叶斯判别法建立了判断城市空气质量等级的判别函数,结果表明将K-均值聚类分析和贝叶斯判别分析用于判别城市空气质量等级的方法准确可靠,能够应用于我国城市空气质量等级预测,最后分析不同等级的城市空气被污染的原因并提出针对性治理建议,为我国城市空气污染治理提供参考.In order to study the urban air quality situation,the sample data of 65 key environmental protection cities in 2018 is analyzed.On the basis of grading the sample key environmental protection cities by using K-mean clustering,the bayes discriminant method is further used to establish the discriminant function to judge the urban air quality grade,the results show that the K-means clustering method and Bayesian discrimination method are accurate and reliable,which can be applied to the prediction of urban air quality grade in China.Finally,the causes of urban air pollution of different levels are analyzed and the corresponding treatment suggestions are put forward,providing reference for the treatment of urban air pollution in China.

关 键 词：空气质量 K-均值聚类分析 贝叶斯判别 等级分类 

分 类 号：X51[环境科学与工程—环境工程] O212.4[理学—概率论与数理统计]