CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: A fresh look at sparse quantile regression Authors:  Paulo Serra - VU Amsterdam (Netherlands) [presenting]
Alexandra Vegelien - VU Amsterdam (Netherlands)
Abstract: In statistics, the aim is often to discover (sometimes impose) structure on observed data, and dimension plays a crucial role in this task. For instance, high-dimensional data sometimes live in a lower-dimensional space, and sparse models are a popular way to represent this. Sparse quantile regression combined with appropriate penalties produces sparse, robust estimators. We will share some results about what kinds of advantages sparse quantile regression brings over mean-based estimators, particularly in terms of robustness, support recovery, correlation between observations and in the design, asymmetric and fat-tailed distributions, and models with quantile level dependent sparseness.