|
Activity Number:
|
21
- JASA T&M Session
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
JASA, Theory and Methods
|
|
Abstract #314405
|
|
|
Title:
|
A Tuning-Free Robust and Efficient Approach to High-Dimensional Regression
|
|
Author(s):
|
Lan Wang* and Bo Peng and Jelena Bradic and Runze Li and Yunan Wu
|
|
Companies:
|
University of Miami and Adobe and UC San Diego and Pennsylvania State University and University of Minnesota
|
|
Keywords:
|
Efficiency;
Heavy-tailed error;
High dimension;
Linear regression;
Tuning parameter;
Robustness
|
|
Abstract:
|
We introduce a new approach for high-dimensional regression. The new procedure overcomes the challenge of tuning parameter selection of Lasso and possesses several appealing properties. It uses an easily simulated tuning parameter that automatically adapts to both the unknown random error distribution and the correlation structure of the design matrix. It is robust with substantial efficiency gain for heavy-tailed random errors while maintaining high efficiency for normal random errors. Comparing with other alternative robust regression procedures, it also enjoys the property of being equivariant when the response variable undergoes a scale transformation. It can be efficiently solved via linear programming. We establish a finite-sample error bound with a near-oracle rate for the new estimator with the simulated tuning parameter. Our results make useful contributions to mending the gap between the practice and theory of Lasso and its variants. We prove that further improvement in efficiency can be achieved by a second-stage enhancement with some light tuning. Our simulations show that the proposed methods often outperform cross-validated Lasso in various settings.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2020 program
|
