The Application Of Quantile Regression In Streaming Data

With the continuous development of Internet technology,the scale and speed of data have presented an explosive growth,more and more data in the form of stream,streaming data is more and more widely used.Recently,as a regression model,quantile regression has been widely used in many fields such as economics,finance and medicine.Compared with ordinary linear regression,quantile regression is more robust to outliers and can better describe different characteristics of data distribution.However,the traditional quantile regression is faced with challenges in parameter estimation of streaming data due to the on-line real-time arrival of streaming data.As an important statistical method for analyzing big data,online updating method can break the storage barrier and computing barrier in some cases by processing the data stream in real time and updating the parameters of the model dynamically.Based on the idea of online updating,this paper proposes a quantile regression method applied to streaming data.The parameter estimation of streaming data is carried out,and the estimator is updated by using the summary statistics of current and historical data.In theory,it is proved that without any additional conditions,the obtained estimators have the same asymptotic distribution as those obtained using all data sets,and the feasibility and accuracy of the proposed method are verified by comparing the simulation experiment and real data with quantile regression using all data sets.