部分线性模型在空气质量指数细颗粒物PM2.5中的分析应用  

Analysis and Application of Partially Linear Models in the Air Quality Index of Fine Particles in PM2.5

作　　者：刘锋[1] 银利 张星[1] 

机构地区：[1]重庆理工大学数学与统计学院,重庆400054

出　　处：《数学的实践与认识》2014年第9期130-134,共5页Mathematics in Practice and Theory

摘　　要：主要是研究细颗粒物PM2.5与其它影响空气质量指数的因素之间的相关性.首先运用主成分分析法对影响细颗粒物PM2.5的五个指标进行降维,然后对降维过后的数据拟合部分线性模型,拟合的效果比一般线性模型与多项式回归模型所拟合的效果更好.This paper is mainly to study the correlation between fine particles of PM2.5 and other factors affecting the air quality index. Firstly, it reduces the dimension of five indices influence the fine particles PM2.5 with the principal component analysis, then, partially linear model is applied to fit the dimensionreduced data, results show that the effect of fitting of partially linear model is better than the general linear model and polynomial regression.

关 键 词：部分线性模型 主成分分析 一般线性模型 

分 类 号：O212.1[理学—概率论与数理统计]