{"id":"arxiv/2603.14991","url":"https://arxiv.org/abs/2603.14991","tags":["math.ST","math.OC"],"score":0,"title":"Wasserstein Distributionally Robust Quantile Regression","domain":"math.ST","source":"arxiv","authors":["Chunxu Zhang","Tiantian Mao","Ruodu Wang"],"reviews":0,"subject":"math.ST","summary":"We study distributionally robust quantile regression using type-$p$ Wasserstein ambiguity sets. We derive a closed-form expression for the worst-case quantile regression loss under general $p$-Wasserstein uncertainty. We further give a uniqueness result showing that for $p>1$, the check loss yields the only class of convex loss functions for which such an additive Wasserstein regularization holds. Our analysis also uncovers qualitative differences between the regimes $p=1$ and $p>1$. When $p>1$, the slope coefficients coincide with those of the regularized formulation, while the intercept…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-16","reactions":{}}