Saved in:
Bibliographic Details
Main Authors: Prokhorov, Boris, Chebykin, Semyon, Gasnikov, Alexander, Beznosikov, Aleksandr
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.01160
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909980937945088
author Prokhorov, Boris
Chebykin, Semyon
Gasnikov, Alexander
Beznosikov, Aleksandr
author_facet Prokhorov, Boris
Chebykin, Semyon
Gasnikov, Alexander
Beznosikov, Aleksandr
contents This paper deals with stochastic optimization problems involving Markovian noise with a zero-order oracle. We present and analyze a novel derivative-free method for solving such problems in strongly convex smooth and non-smooth settings with both one-point and two-point feedback oracles. Using a randomized batching scheme, we show that when mixing time $τ$ of the underlying noise sequence is less than the dimension of the problem $d$, the convergence estimates of our method do not depend on $τ$. This observation provides an efficient way to interact with Markovian stochasticity: instead of invoking the expensive first-order oracle, one should use the zero-order oracle. Finally, we complement our upper bounds with the corresponding lower bounds. This confirms the optimality of our results.
format Preprint
id arxiv_https___arxiv_org_abs_2601_01160
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Gradient-Free Approaches is a Key to an Efficient Interaction with Markovian Stochasticity
Prokhorov, Boris
Chebykin, Semyon
Gasnikov, Alexander
Beznosikov, Aleksandr
Optimization and Control
Machine Learning
This paper deals with stochastic optimization problems involving Markovian noise with a zero-order oracle. We present and analyze a novel derivative-free method for solving such problems in strongly convex smooth and non-smooth settings with both one-point and two-point feedback oracles. Using a randomized batching scheme, we show that when mixing time $τ$ of the underlying noise sequence is less than the dimension of the problem $d$, the convergence estimates of our method do not depend on $τ$. This observation provides an efficient way to interact with Markovian stochasticity: instead of invoking the expensive first-order oracle, one should use the zero-order oracle. Finally, we complement our upper bounds with the corresponding lower bounds. This confirms the optimality of our results.
title Gradient-Free Approaches is a Key to an Efficient Interaction with Markovian Stochasticity
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2601.01160