Font Size: a A A

Some Estimates And Tests Of Panel Data

Posted on:2020-08-19Degree:MasterType:Thesis
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:2417330578971425Subject:Statistics
Abstract/Summary:PDF Full Text Request
In the modern econometric theory system,panel data model is one of the most important components.With the rapid development and maturity of econometric theory,the research on Panel Data Model in China has made a breakthrough development.The first chapter is the introduction,which mainly introduces the background of panel data research and previous research work on panel data.The second chapter studies the composite quantile regression of panel data model.It mainly studies three parts.The first part is the estimation of regression coefficients and the asymptotic normal property of estimates of random effect panel data model obtained by composite quantile regression.The second part mainly studies the asymptotic relative efficiency of panel data compos-ite quantile regression estimation and least square method by calculating the ratio of trace of two estimates of covariance matrix.In the third part,the adaptive lasso penalty composite quantile regression estimation is studied.The idea of adaptive Lasso is used to calculate the estimator of penalty composite quantile regression coefficients,and the asymptotic normality of the estimator is proved.Chapter 3 studies the hypothesis testing of panel model and gives a new test statistic.Through random simulation experiments,it is found that the new test statistic is better than the previous test statistic in controlling the probability of making the first type of error,and the new method is more effective.In the fourth chapter is the practical application.The factors affecting the GDP of Hebei Province are studied by using panel data composite quantile regression method.It is found that the factors are positively related to GDP in Hebei Province,Some suggestions are put forward.
Keywords/Search Tags:panel data, composite quantile regression, asymptotic relative efficiency, asymptotic normality, variable selection, adaptive lasso, hypothesis testing, Monte Carlo
PDF Full Text Request
Related items