Gradient-free algorithm for saddle point problems under overparametrization
Optimization and Control
2024-06-05 v1
Abstract
This paper focuses on solving a stochastic saddle point problem (SPP) under an overparameterized regime for the case, when the gradient computation is impractical. As an intermediate step, we generalize Same-sample Stochastic Extra-gradient algorithm (Gorbunov et al., 2022) to a biased oracle and estimate novel convergence rates. As the result of the paper we introduce an algorithm, which uses gradient approximation instead of a gradient oracle. We also conduct an analysis to find the maximum admissible level of adversarial noise and the optimal number of iterations at which our algorithm can guarantee achieving the desired accuracy.
Cite
@article{arxiv.2406.02308,
title = {Gradient-free algorithm for saddle point problems under overparametrization},
author = {Ekaterina Statkevich and Sofiya Bondar and Darina Dvinskikh and Alexander Gasnikov and Aleksandr Lobanov},
journal= {arXiv preprint arXiv:2406.02308},
year = {2024}
}