EUVIMEDEUVIMED entdecken
Crossref

AGDA+: Proximal Alternating Gradient Descent Ascent Method with a Nonmonotone Adaptive Step-Size Search for Nonconvex Minimax Problems

Xuan Zhang, Qiushui Xu, Necdet Serhat Aybat

SIAM Journal on Optimization · 31.12.2026 · 36 (4) · 2351-2386

DOI: 10.1137/24m1692277 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SIAM Journal on Optimization, online erschienen am 29.10.2026. Verfasst von Xuan Zhang und 2 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Gefördert von: Office of Naval Research.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We consider double-regularized nonconvex–strongly concave minimax problems of the form [Formula: see text], where [Formula: see text] are closed convex and [Formula: see text] is [Formula: see text]-smooth in [Formula: see text] and [Formula: see text]-strongly concave in [Formula: see text]. We propose a proximal alternating gradient descent ascent method AGDA+, which can adaptively choose nonmonotone primal-dual step sizes to compute an approximate stationary point for [Formula: see text] without requiring the knowledge of the global Lipschitz constant [Formula: see text] and the concavity modulus [Formula: see text]. Using a nonmonotone step-size search (backtracking) scheme, AGDA+ stands out for its ability to exploit the local Lipschitz structure and eliminates the need for precise tuning of hyperparameters. AGDA+ achieves the optimal iteration complexity of [Formula: see text], and it is the first step-size search method for nonconvex–strongly concave minimax problems that require only [Formula: see text] calls to [Formula: see text] per backtracking iteration. The numerical experiments demonstrate its robustness and efficiency.

Förderung

  • Office of Naval Research

Zitieren

APA

Zhang, X., Xu, Q., & Aybat, N. S. (2026). AGDA+: Proximal Alternating Gradient Descent Ascent Method with a Nonmonotone Adaptive Step-Size Search for Nonconvex Minimax Problems. SIAM Journal on Optimization, 36(4), 2351-2386. https://doi.org/10.1137/24m1692277

Vancouver

Zhang X, Xu Q, Aybat NS. AGDA+: Proximal Alternating Gradient Descent Ascent Method with a Nonmonotone Adaptive Step-Size Search for Nonconvex Minimax Problems. SIAM Journal on Optimization. 2026;36(4):2351-2386. doi:10.1137/24m1692277

BibTeX
@article{Zhang2026,
  author = {Zhang, Xuan and Xu, Qiushui and Aybat, Necdet Serhat},
  title = {AGDA+: Proximal Alternating Gradient Descent Ascent Method with a Nonmonotone Adaptive Step-Size Search for Nonconvex Minimax Problems},
  journal = {SIAM Journal on Optimization},
  year = {2026},
  volume = {36},
  number = {4},
  pages = {2351--2386},
  doi = {10.1137/24m1692277}
}
Kostenlos starten →