考虑在函数型解释变量部分观测的情况下,用函数线性模型刻画与标量响应变量的关系.基于函数型主成分分析(Functional Principal Component Analysis,简称 FPCA)实现了对缺失部分样本的重构,并通过实证分析,对一组北京市 2010-2014 年间统计的包括部分观测 PM2.5 数值的气象数据,分析了PM2.5 作为部分观测函数型解释变量对标量响应变量平均气温的影响,结果表明了该方法具有处理缺失函数数据的现实意义.
Empirical Analysis of the Relationship between PM2.5 Concentration and Temperature Based on Functional Data Models
Consider the relationship between a scalar response variable and partially observed functional covariates using a functional linear model.By employing Functional Principal Component Analysis(FPCA),we reconstruct the missing parts of the sample data.An empirical analysis is performed on a dataset of meteorological data,including partially observed PM2.5 values,collected in Beijing from 2010 to 2014.This analysis examines the impact of PM2.5,as a partially observed functional covariate,on the scalar response variable of average temperature.The results indicate that this method is practically significant for handling missing functional data.
functional linear modelmissing at randomcomposite quantile regressionPM2.5
NETL
NSTL
万方数据
