Incremental Quasi-Newton Methods with Faster Superlinear Convergence Rates
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (13) · 14097-14105
DOI: 10.1609/aaai.v38i13.29319 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Zhuanghua Liu und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref).
Abstract (Originalsprache)
We consider the finite-sum optimization problem, where each component function is strongly convex and has Lipschitz continuous gradient and Hessian. The recently proposed incremental quasi-Newton method is based on BFGS update and achieves a local superlinear convergence rate that is dependent on the condition number of the problem. This paper proposes a more efficient quasi-Newton method by incorporating the symmetric rank-1 update into the incremental framework, which results in the condition-number-free local superlinear convergence rate. Furthermore, we can boost our method by applying the block update on the Hessian approximation, which leads to an even faster local convergence rate. The numerical experiments show the proposed methods significantly outperform the baseline methods.
Zitieren
APA
Liu, Z., Luo, L., & Low, B. K. H. (2024). Incremental Quasi-Newton Methods with Faster Superlinear Convergence Rates. Proceedings of the AAAI Conference on Artificial Intelligence, 38(13), 14097-14105. https://doi.org/10.1609/aaai.v38i13.29319
Vancouver
Liu Z, Luo L, Low BKH. Incremental Quasi-Newton Methods with Faster Superlinear Convergence Rates. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(13):14097-14105. doi:10.1609/aaai.v38i13.29319
BibTeX
@article{Liu2024,
author = {Liu, Zhuanghua and Luo, Luo and Low, Bryan Kian Hsiang},
title = {Incremental Quasi-Newton Methods with Faster Superlinear Convergence Rates},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {13},
pages = {14097--14105},
doi = {10.1609/aaai.v38i13.29319}
}
EUVIMED