Estimation Of High Conditional Quantiles For Heavy-tailed Distributions

In recent years,in many fields,people tend to pay more attention to the study of the conditional quantile of the tail data,especially the heavy-tailed distribution.In general,effective estimators can be obtained by quantile regression,but traditional quantile regression does not work for high or extremely low quantile estimation.In this paper,combining a new method for estimating the parameter of tail extremum with the traditional quantile regression estimation,we propose an estimation of high conditional quantiles for heavy-tailed distribution(EHH for short).The EHH method can achieve good estimation accuracy and robustness through parameter adjustment.The details of this article are as follows:In the first chapter,a new method for estimating the parameter of the tail extremum is proposed.In the second chapter,we propose the EHH method.In the third chapter,The robustness of the EHH method is studied through data simulation.In the fourth chapter,we show the application effect of EHH method in case analysis.In the fifth chapter,we give the proofs of the theorems in this paper.