Estimation And Application Of Partial Linear Variable Coefficient Quantile Regression Model Based On Data Augmentation Method

Partial linear variable coefficient quantile regression model combines the advantages of linear regression model and nonlinear variable coefficient regression model.Compared with traditional regression model,it has stronger adaptability and better explanatory power,so it is widely used in Clinical trials,biomedical,environmental monitoring and economic research,but the prevalence of censored data has severely affected the validity of the model,and some statistical reasoning methods applied under the complete data set may cause large errors.The existing methods for dealing with censored data are rarely used in partial linear variable coefficient quantile regression models.In this paper,the non-parametric data augmentation method is used to deal with the data censoring problem of partial linear variable coefficient quantile regression model,and then the coefficient estimation and nonlinear partial function estimation of the linear part of the model are obtained.Exploring the performance of this method in partial linear variable coefficient quantile regression model.In the numerical simulation,by setting different types of censoring and different censoring rates,compared with the existing methods,it is found that the estimation efficiency of the method used in this paper is the highest.In the empirical analysis,this paper analyzes the ?-carotene concentration and its influencing factors data and housing prices and its influencing factors in Xindian District,Xinbei City,Taiwan.By constructing a linear variable coefficient quantile regression model,the coefficients are obtained in different quantiles.The estimation at the numerical level reflects the advantages of the research method over other methods by the 95% confidence interval of each coefficient,and explores the relationship between the influence factors on the ?-carotene concentration and the housing price of Xindian District,New Taipei City,and gives Corresponding suggestions.