CMStatistics 2015: Start Registration
View Submission - CMStatistics
B0877
Title: Bootstrapping the de-sparsified Lasso Authors:  Ruben Dezeure - ETH Zurich (Switzerland) [presenting]
Peter Buehlmann - ETH Zurich (Switzerland)
Cun-Hui Zhang - Rutgers University (United States)
Abstract: Assigning statistical significance in high-dimensional linear models has been a very active research area as of late. Some very different approaches have been taken with differing assumptions and empirical performance. We focus on the de-sparsified Lasso, highlight some strengths and weaknesses, and discuss bootstrapping procedures. The implications of model misspecification will be discussed.