Limitations of Deep Learning for Inverse Problems on Digital Hardware
IEEE Transactions on Information Theory · 2023-12 · 69 (12) · 7887-7908
DOI: 10.1109/tit.2023.3326879 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Information Theory, 2023-12. Verfasst von Holger Boche und 2 weiteren Autor:innen. Die Arbeit wurde bisher 28-mal zitiert (Crossref) und nennt 89 Literaturangaben. Gefördert von: German Federal Ministry of Education and Research (BMBF) in the Project Hardware Platforms and Computing Models for Neuromorphic Computing, National Initiative on 6G Communication Systems through the Research Hub 6G-life, Konrad Zuse School of Excellence in Reliable AI [German Academic Exchange Service (DAAD)].
Abstract (Originalsprache)
Deep neural networks have seen tremendous success over the last years. Since the training is performed on digital hardware, in this paper, we analyze what actually can be computed on current hardware platforms modeled as Turing machines, which would lead to inherent restrictions of deep learning. For this, we focus on the class of inverse problems, which, in particular, encompasses any task to reconstruct data from measurements. We prove that finite-dimensional inverse problems are not Banach-Mazur computable for small relaxation parameters. Even more, our results introduce a lower bound on the accuracy that can be obtained algorithmically.
Förderung
- German Federal Ministry of Education and Research (BMBF) in the Project Hardware Platforms and Computing Models for Neuromorphic Computing
- National Initiative on 6G Communication Systems through the Research Hub 6G-life
- Konrad Zuse School of Excellence in Reliable AI [German Academic Exchange Service (DAAD)]
- Munich Center for Machine Learning
- German Research Foundation
- LMUexcellent
- BMBF, by the Free State of Bavaria under the Excellence Strategy of the Federal Government and the Länder and by the Hightech Agenda Bavaria
Zitieren
APA
Boche, H., Fono, A., & Kutyniok, G. (2023). Limitations of Deep Learning for Inverse Problems on Digital Hardware. IEEE Transactions on Information Theory, 69(12), 7887-7908. https://doi.org/10.1109/tit.2023.3326879
Vancouver
Boche H, Fono A, Kutyniok G. Limitations of Deep Learning for Inverse Problems on Digital Hardware. IEEE Transactions on Information Theory. 2023;69(12):7887-7908. doi:10.1109/tit.2023.3326879
BibTeX
@article{Boche2023,
author = {Boche, Holger and Fono, Adalbert and Kutyniok, Gitta},
title = {Limitations of Deep Learning for Inverse Problems on Digital Hardware},
journal = {IEEE Transactions on Information Theory},
year = {2023},
volume = {69},
number = {12},
pages = {7887--7908},
doi = {10.1109/tit.2023.3326879}
}
EUVIMED