Regularized regression with the l1 penalty is a popular approach for variable selection and coefficient estimation. For a unified treatment of the l1-constrained model selection, Wang and Leng (2007) proposed the least squares approximation method (LSA) for a fixed dimension. LSA makes use of a...

We consider how to incorporate auxiliary information to improve quantile regression via empirical likelihood. We propose a novel framework and show that our approach yields more efficient estimates compared to those from the conventional quantile regression. The efficiency gain is quantified...