A Note On Distributed Quantile Regression By Pilot Sampling And One-step Updating

Abstract

Quantile regression is amethod of fundamental importance. Howto efficiently conduct quantile regression for a large dataset on a distributed systemis of great importance.We showthat the popularly used one-shot estimation is statistically inefficient if data are not randomly distributed across different workers. To fix the problem, a novel one-step estimation method is developed with the following nice properties. First, the algorithm is communication efficient. That is the communication cost demanded is practically acceptable. Second, the resulting estimator is statistically efficient. That is its asymptotic covariance is the same as that of the global estimator. Third, the estimator is robust against data distribution. That is its consistency is guaranteed even if data are not randomly distributed across different workers. Numerical experiments are provided to corroborate our findings. A real example is also presented for illustration.