Carath\'eodory Theory and A Priori Estimates for Continuity Inclusions in the Space of Probability Measures
Beno\^it Bonnet-Weill, H\'el\`ene Frankowska

TL;DR
This paper extends the theory of differential inclusions to Wasserstein spaces of probability measures, proving a new existence result under Carathéodory conditions and improving solution estimates using a set-valued Euler scheme.
Contribution
It introduces a novel existence theorem for continuity inclusions in Wasserstein spaces under minimal regularity assumptions, utilizing a set-valued Euler scheme adapted from Filippov's method.
Findings
Established a Peano-type existence result for continuity inclusions.
Developed new estimates and compactness properties for solutions.
Enhanced the understanding of solution set relaxation in the Wasserstein space.
Abstract
In this article, we extend the foundations of the theory of differential inclusions in the space of compactly supported probability measures, introduced recently by the authors, to the setting of general Wasserstein spaces. In this context, we prove a novel existence result \`a la Peano for this class of dynamics under mere Carath\'eodory regularity assumptions. The latter is based on a natural, yet previously unexplored set-valued adaptation of the semi-discrete Euler scheme proposed by Filippov to study ordinary differential equations whose right-hand sides are measurable in the time variable. By leveraging some of the underlying methods along with new estimates for solutions of continuity equations, we also bring substantial improvements to the earlier versions of the Filippov estimates, compactness and relaxation properties of the solution sets of continuity inclusions, which are…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAnomaly Detection Techniques and Applications · Geometric Analysis and Curvature Flows
