Abstract
We study the problem of reconstructing solutions of inverse problems when only noisy measurements are available. We assume that the problem can be modeled with an infinite-dimensional forward operator that is not continuously invertible. Then, we restrict this forward operator to finite-dimensional spaces so that the inverse is Lipschitz continuous. For the inverse operator, we demonstrate that there exists a neural network which is a robust-to-noise approximation of the operator. In addition, we show that these neural networks can be learned from appropriately perturbed training data. We demonstrate the admissibility of this approach to a wide range of inverse problems of practical interest. Numerical examples are given that support the theoretical findings.
Originalsprache | Englisch |
---|---|
Seiten (von - bis) | 49-91 |
Seitenumfang | 43 |
Fachzeitschrift | Analysis and Applications |
Jahrgang | 21 |
Ausgabenummer | 1 |
DOIs | |
Publikationsstatus | Veröffentlicht - 1 Jan. 2023 |
ÖFOS 2012
- 102018 Künstliche Neuronale Netze