Partial Functional Linear Regression Model Based on Beijing PM2.5 Data Set

In view of the problem that missing PM2.5 data in Beijing, a partial functional linear regression model with missing data was proposed to fit the data to explore how wind speed and temperature affect PM2.5 concentration. By using the method of functional principal component analysis, the estimator of the bivariate parameter function is obtained, and the empirical logarithm likelihood ratio statistics of the linear partial parameters are constructed by using the inverse probability weighting method. The index function R(t) was used to access the fitting results of the data set, and it was concluded that the partial functional linear regression model with missing data had a better fitting effect on the data set, could deal with PM2.5 missing data more flexibly and accurately, and could specifically explain the practical significance of the data.