Quantile Regression Method And Empirical Study Of Panel Data Based On Elastic Net

In recent years,panel data has been widely used in various fields of life.Variable selection is an important issue in the establishment of statistical process models,but there are some shortcomings in the traditional stepwise regression method combined with information criteria.In 1996,Tibshirani.R proposed a variable selection method called Lasso,which can overcome the traditional method.In2005,Zou and Hastie proposed an improved method called Elastic net.This paper describes the quantile regression method of panel data based on elastic net,and carries out empirical research by using the method.Firstly,a systematic comparative study of Lasso and elastic net based on panel data model is summarized,which shows the relationship between them.Then,the research of elastic net method based on quantile model of panel data is analyzed,and the data simulation is used to verify the validity of the method.The elastic net is analyzed and compared.Finally,the quantitative regression of the elastic net is used to make an empirical analysis of the actual carbon emission data.The research framework of this paper is as follows: Chapter 1 is the introduction part.It describes significance of the subject and the research background briefly.Then it summarizes the subject to the present research,and explains the main idea of this paper.Chapter 2 is the theoretical elaboration of relevant knowledge.On the one hand,it introduces the panel data and the balanced panel data model,on the other hand,it also introduces the evaluation criteria of the model.Thirdly,it introduces the evaluation criteria of the model.Chapter 4 studies the elastic net method based on the panel data quantile model,and shows its group effect by theorem.Chapter 5 carries out numerical simulation on the basis of theory to further verify the feasibility of the method.Chapter 6 applies carbon emissions.The final chapter is the outlook and summary of this paper.