PaperRegret-Optimal Wasserstein-Robust Linear Regression
Regret-Optimal Wasserstein-Robust Linear Regression
A research paper examining the suboptimality of distributionally robust optimization (DRO) in multivariate linear regression. It proposes two alternatives, minimax regret optimization (MRO) and spectral norm regularization (SNR), and proves that they attain minimax-optimal rates.
Timeline (last 90 days)
Nov 2
The paper "Regret-Optimal Wasserstein-Robust Linear Regression" proposes two methods, minimax regret optimization (MRO) and spectral norm regularization (SNR), and claims they can achieve minimax optimal rates under broader conditions.
Unverified50%