Logo image
Quantifying Distributional Model Risk in Marginal Problems via Optimal Transport
Preprint   Open access

Quantifying Distributional Model Risk in Marginal Problems via Optimal Transport

Yanqin Fan, Hyeonseok Park and Gaoqian Xu
arXiv
arXiv
07/03/2023
DOI: 10.48550/arxiv.2307.00779
url
https://doi.org/10.1287/moor.2024.0557View
Published (Version of record) This article has now been published in a journal and has been peer-reviewed by subject experts. This version may differ significantly from the preprint version. Access restricted to faculty, staff and students
url
https://doi.org/10.48550/arxiv.2307.00779View
Preprint (Author's original) This preprint has not been evaluated by subject experts through peer review. Preprints may undergo extensive changes and/or become peer-reviewed journal articles. Open Access

Abstract

This paper studies distributional model risk in marginal problems, where each marginal measure is assumed to lie in a Wasserstein ball centered at a fixed reference measure with a given radius. Theoretically, we establish several fundamental results including strong duality, finiteness of the proposed Wasserstein distributional model risk, and the existence of an optimizer at each radius. In addition, we show continuity of the Wasserstein distributional model risk as a function of the radius. Using strong duality, we extend the well-known Makarov bounds for the distribution function of the sum of two random variables with given marginals to Wasserstein distributionally robust Markarov bounds. Practically, we illustrate our results on four distinct applications when the sample information comes from multiple data sources and only some marginal reference measures are identified. They are: partial identification of treatment effects; externally valid treatment choice via robust welfare functions; Wasserstein distributionally robust estimation under data combination; and evaluation of the worst aggregate risk measures.
Mathematics - Optimization and Control Mathematics - Statistics Theory Statistics - Theory

Details

Metrics

1 Record Views
Logo image