Optimal Reach Estimation and Metric Learning
hal.structure.identifier | Laboratoire de Probabilités, Statistique et Modélisation [LPSM (UMR_8001)] | |
dc.contributor.author | Aamari, Eddie
HAL ID: 181405 ORCID: 0000-0003-4516-3009 | |
hal.structure.identifier | CEntre de REcherches en MAthématiques de la DEcision [CEREMADE] | |
dc.contributor.author | Berenfeld, Clément | |
hal.structure.identifier | Laboratoire de Probabilités, Statistique et Modélisation [LPSM (UMR_8001)] | |
dc.contributor.author | Levrard, Clément | |
dc.date.accessioned | 2022-12-09T11:28:39Z | |
dc.date.available | 2022-12-09T11:28:39Z | |
dc.date.issued | 2022 | |
dc.identifier.uri | https://basepub.dauphine.psl.eu/handle/123456789/23324 | |
dc.language.iso | en | en |
dc.subject.ddc | 519 | en |
dc.title | Optimal Reach Estimation and Metric Learning | en |
dc.type | Document de travail / Working paper | |
dc.description.abstracten | We study the estimation of the reach, an ubiquitous regularity parameter in manifold estimation and geometric data analysis. Given an i.i.d. sample over an unknown d-dimensional Ck-smooth submanifold of RD , we provide optimal nonasymptotic bounds for the estimation of its reach. We build upon a formulation of the reach in terms of maximal curvature on one hand, and geodesic metric distortion on the other hand. The derived rates are adaptive, with rates depending on whether the reach of M arises from curvature or from a bottleneck structure. In the process, we derive optimal geodesic metric estimation bounds. | en |
dc.publisher.city | Paris | en |
dc.identifier.citationpages | 47 | en |
dc.relation.ispartofseriestitle | Cahier de recherche CEREMADE, Université Paris Dauphine-PSL | en |
dc.subject.ddclabel | Probabilités et mathématiques appliquées | en |
dc.identifier.citationdate | 2022 | |
dc.description.ssrncandidate | non | |
dc.description.halcandidate | non | en |
dc.description.readership | recherche | en |
dc.description.audience | International | en |
dc.date.updated | 2022-12-09T11:23:15Z | |
hal.author.function | aut | |
hal.author.function | aut | |
hal.author.function | aut |