Shuffling Gradient-Based Methods for Nonconvex-Concave Minimax Optimization

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tran-Dinh, Quoc, Tran, Trang H., Nguyen, Lam M.
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910677539487744
author Tran-Dinh, Quoc
Tran, Trang H.
Nguyen, Lam M.
author_facet Tran-Dinh, Quoc
Tran, Trang H.
Nguyen, Lam M.
contents This paper aims at developing novel shuffling gradient-based methods for tackling two classes of minimax problems: nonconvex-linear and nonconvex-strongly concave settings. The first algorithm addresses the nonconvex-linear minimax model and achieves the state-of-the-art oracle complexity typically observed in nonconvex optimization. It also employs a new shuffling estimator for the "hyper-gradient", departing from standard shuffling techniques in optimization. The second method consists of two variants: semi-shuffling and full-shuffling schemes. These variants tackle the nonconvex-strongly concave minimax setting. We establish their oracle complexity bounds under standard assumptions, which, to our best knowledge, are the best-known for this specific setting. Numerical examples demonstrate the performance of our algorithms and compare them with two other methods. Our results show that the new methods achieve comparable performance with SGD, supporting the potential of incorporating shuffling strategies into minimax algorithms.
format Preprint
id arxiv_https___arxiv_org_abs_2410_22297
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Shuffling Gradient-Based Methods for Nonconvex-Concave Minimax Optimization
Tran-Dinh, Quoc
Tran, Trang H.
Nguyen, Lam M.
Optimization and Control
Machine Learning
This paper aims at developing novel shuffling gradient-based methods for tackling two classes of minimax problems: nonconvex-linear and nonconvex-strongly concave settings. The first algorithm addresses the nonconvex-linear minimax model and achieves the state-of-the-art oracle complexity typically observed in nonconvex optimization. It also employs a new shuffling estimator for the "hyper-gradient", departing from standard shuffling techniques in optimization. The second method consists of two variants: semi-shuffling and full-shuffling schemes. These variants tackle the nonconvex-strongly concave minimax setting. We establish their oracle complexity bounds under standard assumptions, which, to our best knowledge, are the best-known for this specific setting. Numerical examples demonstrate the performance of our algorithms and compare them with two other methods. Our results show that the new methods achieve comparable performance with SGD, supporting the potential of incorporating shuffling strategies into minimax algorithms.
title Shuffling Gradient-Based Methods for Nonconvex-Concave Minimax Optimization
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2410.22297