Paperlane

no. 5 · Friday 30 October 2026

Wasserstein Distributionally Robust Online Learning

Guixian Chen, Salar Fattahi and Soroosh Shafiee

math.OC · v3

Abstract

We study online Wasserstein distributionally robust optimization (DRO), where the ambiguity set is a Wasserstein ball centered at the empirical distribution and evolves with each observation. We formulate each round as a zero-sum game and develop an online distributional best-response algorithm that requires only one projected subgradient step per iteration. Combining regret analysis with measure concentration, we establish explicit, instance-dependent, and high-probability convergence rates to the optimal offline DRO value.

fields
Optimization and Control
pdf ↗ page arXiv:2602.20403 ↗