The Proximal Bootstrap for Finite-Dimensional Regularized Estimators
AEA Papers and Proceedings
vol. 111,
May 2021
(pp. 616-20)
Abstract
We propose a proximal bootstrap that can consistently estimate the limiting distribution of √n consistent estimators with nonstandard asymptotic distributions in a computationally efficient manner by formulating the proximal bootstrap estimator as the solution to a convex optimization problem, which can have a closed-form solution for certain designs. This paper considers the application to finite-dimensional regularized estimators, such as the Lasso, ℓ1-norm regularized quantile regression, ℓ1-norm support vector regression, and trace regression via nuclear norm regularization.Citation
Li, Jessie. 2021. "The Proximal Bootstrap for Finite-Dimensional Regularized Estimators." AEA Papers and Proceedings, 111: 616-20. DOI: 10.1257/pandp.20211036Additional Materials
JEL Classification
- C51 Model Construction and Estimation
- C15 Statistical Simulation Methods: General