A0382
Title: Composite smoothed quantile regression
Authors: Xiaozhou Wang - East China Normal University (China) [presenting]
Abstract: Composite quantile regression (CQR) is an efficient method to estimate the parameters of the linear model with non-Gaussian random noise. The non-smoothness of CQR loss prevents many efficient methods from being used. The composite smoothed quantile regression model is proposed, and the inference problem is investigated for a large-scale dataset. The algorithm and theoretical properties are given. Extensive numerical experiments on both simulated and real data are conducted to demonstrate the good performance of the proposed estimator compared to some baselines.
