Adaptive two-points zero-order accelerated methods for stochastic convex optimization problems

Yuriy Dorn, Osman Osmanov, Timofei Loginov, Nazarii Tupitsa, Darina Dvinskikh, Alexander Gasnikov

arXivarxiv/2610.15202Oct 10, 2026math.OCscore 0

We study stochastic convex optimization, where each sampled function can be evaluated at two arbitrary points. Our accelerated method chooses step sizes, batch sizes and smoothing radii from function values. A forward difference between smoothing scales controls the radius; we account for changes in the smoothed objective. Under the stated regularity and second-moment assumptions, exact function evaluations give \begin{multline*} \mathbb{E}\bigl[f(x_N)-f(x^\star)\bigr] \le \widetilde{O}\biggl( \frac{\sqrt{d}\,M_2 D}{\sqrt{T}} +\min\biggl( \frac{L D^2}{N^2},\; \frac{d^{1/4} M…

SourceJSON

Comments and reviews

No comments yet.

API

curl /api/v1/records/arxiv/2610.15202
curl -X POST /api/v1/records/arxiv/2610.15202/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2610.15202/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'