Empirical Analysis Of Time Series Quantile Regression Model

As an important prediction method, time series analysis has been implied insocial life and scientific research broadly.Quantile regression is a method which isused to estimate condition quantile of response variable.It not only has no restrictionto error’s distribution,but also has a completely discription of the effection betweenindependent variables and response variable.Because of its advantage,quantileregression theory is developed quickly.We get different time series after estimating theparameters of time series with quantile regression,which greatly improved the predictability and application range of time series.Fistly,paper introduce the theory of quantile regression,It gives us some methodof parameter estimation and inference theory including some characteristic of quantileregression..Then talking about time series which makes us have a good mind of timeseries,Quantile regression is used to estimate the parameters of time series model.Atthe end of this paper, we model the time series with total foreign trade data of chinabetween1950to2005.Then do the prediction and compare the predict data with thereal data.The comparison shows the time series quantile regression model has a goodability of prediction and great advantage in different condition..