EUVIMEDEUVIMED entdecken
Crossref

Implicit Balancing and Regularization: Generalization and Convergence Guarantees for Overparameterized Asymmetric Matrix Sensing

Mahdi Soltanolkotabi, Dominik Stöger, Changzhi Xie

IEEE Transactions on Information Theory · 2025-04 · 71 (4) · 2991-3037

DOI: 10.1109/tit.2025.3530335 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Information Theory, 2025-04. Verfasst von Mahdi Soltanolkotabi und 2 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 81 Literaturangaben. Gefördert von: Packard Fellowship in Science and Engineering, Sloan Fellowship in Mathematics, NSF-CAREER.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Kein Abstract verfügbar.

Förderung

  • Packard Fellowship in Science and Engineering
  • Sloan Fellowship in Mathematics
  • NSF-CAREER
  • Defense Advanced Research Projects Agency (DARPA) Learning with Less Labels (LwLL) and Fast Network Interface Cards (FastNICS) Programs
  • NSF-Civil Infrastructure Systems

Zitieren

APA

Soltanolkotabi, M., Stöger, D., & Xie, C. (2025). Implicit Balancing and Regularization: Generalization and Convergence Guarantees for Overparameterized Asymmetric Matrix Sensing. IEEE Transactions on Information Theory, 71(4), 2991-3037. https://doi.org/10.1109/tit.2025.3530335

Vancouver

Soltanolkotabi M, Stöger D, Xie C. Implicit Balancing and Regularization: Generalization and Convergence Guarantees for Overparameterized Asymmetric Matrix Sensing. IEEE Transactions on Information Theory. 2025;71(4):2991-3037. doi:10.1109/tit.2025.3530335

BibTeX
@article{Soltanolkotabi2025,
  author = {Soltanolkotabi, Mahdi and Stöger, Dominik and Xie, Changzhi},
  title = {Implicit Balancing and Regularization: Generalization and Convergence Guarantees for Overparameterized Asymmetric Matrix Sensing},
  journal = {IEEE Transactions on Information Theory},
  year = {2025},
  volume = {71},
  number = {4},
  pages = {2991--3037},
  doi = {10.1109/tit.2025.3530335}
}
Kostenlos starten →