Transformed Central Quantile Subspace

Abstract: Quantile regression (QR) is becoming increasingly popular due to its relevance in many scientific investigations. However, application of QR can become very challenging when dealing with high-dimensional data, making it necessary to use dimension reduction techniques. Existing dimension reduction techniques focus on the entire conditional distribution. We turn our attention to dimension reduction techniques for conditional quantiles and introduce a method that serves as an intermediate step between linear and nonlinear dimension reduction. The idea is to apply existing linear dimension reduction techniques on the transformed predictors. The proposed estimator, which is shown to be root-n consistent, is demonstrated through simulation examples and real data applications. Our results suggest that this method outperforms linear dimension reduction for conditional quantiles.