New perspectives on regularization and computation in optimal transport-based distributionally robust optimization
We study optimal transport-based distributionally robust optimization problems where a
fictitious adversary, often envisioned as nature, can choose the distribution of the uncertain …
fictitious adversary, often envisioned as nature, can choose the distribution of the uncertain …
Superquantiles at work: Machine learning applications and efficient subgradient computation
R. Tyrell Rockafellar and his collaborators introduced, in a series of works, new regression
modeling methods based on the notion of superquantile (or conditional value-at-risk). These …
modeling methods based on the notion of superquantile (or conditional value-at-risk). These …
Distributionally robust optimization with bias and variance reduction
We consider the distributionally robust optimization (DRO) problem with spectral risk-based
uncertainty set and $ f $-divergence penalty. This formulation includes common risk …
uncertainty set and $ f $-divergence penalty. This formulation includes common risk …
Risk-adaptive approaches to learning and decision making: A survey
JO Royset - arXiv preprint arXiv:2212.00856, 2022 - arxiv.org
Uncertainty is prevalent in engineering design, statistical learning, and decision making
broadly. Due to inherent risk-averseness and ambiguity about assumptions, it is common to …
broadly. Due to inherent risk-averseness and ambiguity about assumptions, it is common to …
Sensitivity analysis of Wasserstein distributionally robust optimization problems
We consider sensitivity of a generic stochastic optimization problem to model uncertainty.
We take a non-parametric approach and capture model uncertainty using Wasserstein balls …
We take a non-parametric approach and capture model uncertainty using Wasserstein balls …
Distributionally robust learning with weakly convex losses: Convergence rates and finite-sample guarantees
L Zhu, M Gürbüzbalaban, A Ruszczyński - arXiv preprint arXiv:2301.06619, 2023 - arxiv.org
We consider a distributionally robust stochastic optimization problem and formulate it as a
stochastic two-level composition optimization problem with the use of the mean …
stochastic two-level composition optimization problem with the use of the mean …
On approximations of data-driven chance constrained programs over Wasserstein balls
Distributionally robust chance constrained programs minimize a deterministic cost function
subject to the satisfaction of one or more safety conditions with high probability, given that …
subject to the satisfaction of one or more safety conditions with high probability, given that …
Coordinate linear variance reduction for generalized linear programming
We study a class of generalized linear programs (GLP) in a large-scale setting, which
includes simple, possibly nonsmooth convex regularizer and simple convex set constraints …
includes simple, possibly nonsmooth convex regularizer and simple convex set constraints …
Distributional uncertainty propagation via optimal transport
This paper addresses the limitations of standard uncertainty models, eg, robust (norm-
bounded) and stochastic (one fixed distribution, eg, Gaussian), and proposes to model …
bounded) and stochastic (one fixed distribution, eg, Gaussian), and proposes to model …
The performance of Wasserstein distributionally robust M-estimators in high dimensions
L Aolaritei, S Shafieezadeh-Abadeh… - arXiv preprint arXiv …, 2022 - arxiv.org
Wasserstein distributionally robust optimization has recently emerged as a powerful
framework for robust estimation, enjoying good out-of-sample performance guarantees, well …
framework for robust estimation, enjoying good out-of-sample performance guarantees, well …