An Improved Interior Point Algorithm for Quantile Regression

Abstract

Quantile regression is a powerful statistical technique for estimating the quantiles of a conditional distribution on the values of covariates. It has been widely used in many fields. In this paper, an improved interior point algorithm for quantile regression is proposed. The algorithm introduces multiple centrality corrections technique into the interior point algorithm for quantile regression. The purpose of introducing the multiple centrality corrections technique is to reduce the overall solution time required to solve a quantile regression problem. The computational experiments results constitute evidence of the improvement obtained with the use of multiple centrality correction technique combined with the interior point algorithm. INDEX TERMS quantile regression, multiple centrality corrections, interior point algorithm.