Convergence bounds for empirical nonlinear least-squares
dc.bibliographicCitation.firstPage | 79 | eng |
dc.bibliographicCitation.issue | 1 | eng |
dc.bibliographicCitation.journalTitle | Mathematical modelling and numerical analysis : an international journal on applied mathematics | eng |
dc.bibliographicCitation.lastPage | 104 | eng |
dc.bibliographicCitation.volume | 56 | eng |
dc.contributor.author | Eigel, Martin | |
dc.contributor.author | Schneider, Reinhold | |
dc.contributor.author | Trunschke, Philipp | |
dc.date.accessioned | 2022-06-16T11:44:02Z | |
dc.date.available | 2022-06-16T11:44:02Z | |
dc.date.issued | 2022 | |
dc.description.abstract | We consider best approximation problems in a nonlinear subset ā³ of a Banach space of functions (š±,ā„ā¢ā„). The norm is assumed to be a generalization of the L 2-norm for which only a weighted Monte Carlo estimate ā„ā¢ā„n can be computed. The objective is to obtain an approximation vāāāā³ of an unknown function uāāāš± by minimizing the empirical norm ā„uā āā vā„n. We consider this problem for general nonlinear subsets and establish error bounds for the empirical best approximation error. Our results are based on a restricted isometry property (RIP) which holds in probability and is independent of the specified nonlinear least squares setting. Several model classes are examined and the analytical statements about the RIP are compared to existing sample complexity bounds from the literature. We find that for well-studied model classes our general bound is weaker but exhibits many of the same properties as these specialized bounds. Notably, we demonstrate the advantage of an optimal sampling density (as known for linear spaces) for sets of functions with sparse representations. | eng |
dc.description.version | publishedVersion | eng |
dc.identifier.uri | https://oa.tib.eu/renate/handle/123456789/9056 | |
dc.identifier.uri | https://doi.org/10.34657/8094 | |
dc.language.iso | eng | eng |
dc.publisher | Les Ulis : EDP Sciences | eng |
dc.relation.doi | https://doi.org/10.1051/m2an/2021070 | |
dc.relation.essn | 2804-7214 | |
dc.rights.license | CC BY 4.0 Unported | eng |
dc.rights.uri | https://creativecommons.org/licenses/by/4.0/ | eng |
dc.subject.ddc | 510 | eng |
dc.subject.other | Convergence rates | eng |
dc.subject.other | Error analysis | eng |
dc.subject.other | Tensor networks | eng |
dc.subject.other | Weighted nonlinear least squares | eng |
dc.subject.other | Weighted sparsity | eng |
dc.title | Convergence bounds for empirical nonlinear least-squares | eng |
dc.type | Article | eng |
dc.type | Text | eng |
tib.accessRights | openAccess | eng |
wgl.contributor | WIAS | eng |
wgl.subject | Mathematik | eng |
wgl.type | Zeitschriftenartikel | eng |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- Convergence_bounds_for_empirical_nonlinear_least-squares.pdf
- Size:
- 1.52 MB
- Format:
- Adobe Portable Document Format
- Description: