Robust Decentralized Learning with Local Updates and Gradient Tracking

2024-05-02

arXiv.org

AI Summary

This work addresses the dual challenges of statistical heterogeneity and adversarial robustness in distributed learning settings—particularly federated learning, IoT, and edge computing. We propose Dec-FedTrack, the first algorithm that jointly integrates local updates and decentralized gradient tracking within a nonconvex-strongly-concave minimax optimization framework, enabling robust decentralized adversarial training over graph-structured networks. Theoretically, we establish convergence to stationary points under realistic assumptions, with provable communication efficiency, convergence guarantees under data heterogeneity, and certified adversarial robustness. Empirically, Dec-FedTrack consistently outperforms existing decentralized methods across heterogeneous data distributions and diverse adversarial attacks—including PGD, FGSM, and AutoAttack—demonstrating superior generalization and robustness without requiring centralized coordination or server-side aggregation.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Decentralized Web and Fediverse systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

As distributed learning applications such as Federated Learning, the Internet of Things (IoT), and Edge Computing grow, it is critical to address the shortcomings of such technologies from a theoretical perspective. As an abstraction, we consider decentralized learning over a network of communicating clients or nodes and tackle two major challenges: data heterogeneity and adversarial robustness. We propose a decentralized minimax optimization method that employs two important modules: local updates and gradient tracking. Minimax optimization is the key tool to enable adversarial training for ensuring robustness. Having local updates is essential in Federated Learning (FL) applications to mitigate the communication bottleneck, and utilizing gradient tracking is essential to proving convergence in the case of data heterogeneity. We analyze the performance of the proposed algorithm, Dec-FedTrack, in the case of nonconvex-strongly concave minimax optimization, and prove that it converges a stationary point. We also conduct numerical experiments to support our theoretical findings.

Problem

• Address data heterogeneity in decentralized learning networks.
• Ensure adversarial robustness in distributed learning applications.
• Mitigate communication bottlenecks with local updates and gradient tracking.

Innovation

• Decentralized minimax optimization for robustness
• Local updates to reduce communication overhead
• Gradient tracking ensures convergence with data heterogeneity