arXiv · 2611.15938
Wasserstein Stability of the Nerve Map
Abstract
Given a locally finite open cover of a paracompact space, a subordinate partition of unity defines a map into the geometric realization of the nerve. When the cover is good, the nerve theorem asserts that this map is a homotopy equivalence. In applied topology, one specializes this construction to covers of metric data by open balls centered at finitely many landmarks. The corresponding nerve map represents the data inside the Čech complex, enabling one to pull cohomological or spectral invariants back to the data. Since point cloud data often comes equipped with weights encoding domain-specific importance, we introduce a weighted version of the nerve map and study its stability under perturbations of the landmarks and weights using metrics from optimal transport. Our main result is a Lipschitz stability theorem for the weighted nerve map under a uniform nondegeneracy assumption. We state the bound in terms of a scale-dependent optimal transport distance, and give an example where our estimate is attained. Since this distance is dominated by a multiple of the $1$-Wasserstein distance, Wasserstein stability follows as a corollary. We also give geometric conditions ensuring nondegeneracy, obtain explicit estimates of the Lipschitz constant for closed Riemannian manifolds, and sketch applications to Laplacian eigenvector cascading and Eilenberg--MacLane coordinates.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Kenneth McCabe, Jose A. Perea. 2026-11-11. Wasserstein Stability of the Nerve Map. https://searcharxiv.com/abs/2611.15938
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Searcharxiv by Indiciae: BibTeX · RIS.