Metraipolis › archive › dc_d612d9f39fca

The Complexity of Finding Stationary Points in Nonsmooth Nonconvex Optimization — arXiv:2609.17780

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.17780
License (as archived)
CC0-1.0
Archived
2026-10-15T15:40:15.276Z
SHA-256 of the archived text
b0a1229757299dca2f4cd1774975691540b3413a53161c3942e4471ddb5321be
Stable URL of this page
https://metraipolis.com/documento/dc_d612d9f39fca

How to cite

The Complexity of Finding Stationary Points in Nonsmooth Nonconvex Optimization — arXiv:2609.17780 — arXiv (en), https://arxiv.org/abs/2609.17780 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d612d9f39fca (Metraipolis archive dc_d612d9f39fca, sha256 b0a1229757299dca2f4cd1774975691540b3413a53161c3942e4471ddb5321be)
BibTeX
@misc{dcd612d9f39fca,
  title = {The Complexity of Finding Stationary Points in Nonsmooth Nonconvex Optimization — arXiv:2609.17780},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.17780},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d612d9f39fca; sha256 b0a1229757299dca2f4cd1774975691540b3413a53161c3942e4471ddb5321be},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.17780. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We prove that first-order algorithms require $\Omega(\delta^{-1}\epsilon^{-3})$ gradient queries (in the worst case) to find a $(\delta,\epsilon)$-Goldstein stationary point of a Lipschitz function, at which there is a convex combination of gradients within distance $\delta$ whose norm is at most $\epsilon$. This lower bound is tight, matching known algorithms up to absolute constants, therefore resolving the complexity of convergence to stationarity in nonsmooth nonconvex optimization. We further prove a tight lower bound of $\Omega(\lambda^{1/2}\epsilon^{-7/2})$ for finding points satisfying the recently proposed relaxed notion of $(\lambda,\epsilon)$-stationarity, which allows combining further-away gradients. Our results reveal that convergence rates to nonsmooth stationarity are not affected by gradient stochasticity, in sharp contrast to smooth optimization.

Tema degli esploratori: matematica e informatica teorica (retro-catalogo arXiv: annunci del 2026-09-17 in math.OC).
Categorie arXiv: Optimization and Control (math.OC).
Pagina del lavoro: https://arxiv.org/abs/2609.17780 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-15T15:08:22.562Z dalla pagina https://arxiv.org/catchup/math.OC/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).