Federated Minimax Optimization

Federated minimax optimization addresses the challenge of training models with adversarial or constrained objectives across decentralized datasets, preserving data privacy while mitigating issues like data heterogeneity and fairness. Current research focuses on developing communication-efficient algorithms, such as those employing gradient tracking or smoothing techniques, and analyzing their convergence rates under various non-convex settings, including nonconvex-strongly-concave and nonconvex-nonconcave scenarios. These advancements are crucial for improving the scalability and robustness of federated learning in applications like GAN training and fair classification, where minimax formulations are prevalent. The ultimate goal is to achieve efficient and provably convergent algorithms for solving these complex distributed optimization problems.

Papers

May 7, 2024

Fast Decentralized Gradient Tracking for Federated Minimax Optimization with Local Updates
Chris Junchi Li
Minimax Optimization Local Context Gradient Tracking Decentralized Gradient Federated Minimax Optimization

November 2, 2023

Stochastic Smoothed Gradient Descent Ascent for Federated Minimax Optimization
Wei Shen, Minhui Huang, Jiawei Zhang, Cong Shen
Minimax Optimization Gradient Descent Ascent Federated Stochastic Federated Minimax Optimization

July 10, 2023

Fairness-aware Federated Minimax Optimization with Convergence Guarantee
Gerry Windiarto Mohamad Dunda, Shenghui Song
Procedural Fairness Group Fairness Convergence Guarantee Fairness Constraint Lagrangian Relaxation Federated Minimax Optimization

February 8, 2023

Federated Minimax Optimization with Client Heterogeneity
Pranay Sharma, Rohan Panda, Gauri Joshi
General Nonconvex Minimax Optimization Client Heterogeneity Nonconvex Nonconcave Problem Federated Minimax Optimization

November 14, 2022

Adaptive Federated Minimax Optimization with Lower Complexities
Feihu Huang, Xinrui Wang, Junyi Li, Songcan Chen
Low Complexity Minimax Problem Federated Minimax Optimization Minimax Optimal Algorithm Nonconvex Minimax Optimization

October 2, 2022

SAGDA: Achieving $\mathcal{O}(\epsilon^{-2})$ Communication Complexity in Federated Min-Max Learning
Haibo Yang, Zhuqing Liu, Xin Zhang, Jia Liu
Communication Complexity Gradient Descent Ascent Stochastic Gradient Descent Ascent Federated Minimax Optimization

June 2, 2022

A Communication-efficient Algorithm with Linear Convergence for Federated Minimax Learning
Zhenyu Sun, Ermin Wei
Communication Efficient Linear Convergence Gradient Descent Ascent Minimax Q Learning Federated Minimax Optimization Minimax Optimal Algorithm

March 9, 2022