用博弈算法提升联邦学习公平性，让不同数据的客户端表现更均衡。
FedMABA: Towards Fair Federated Learning through Multi-Armed Bandits Allocation
- 引入多臂老虎机机制优化客户端分配策略
- 在非独立同分布场景下显著降低性能差异
- 适合关注模型公平性的研究者与应用开发者
数据隐私日益受重视，推动了联邦学习（FL）的发展。然而，客户端间的数据统计异质性导致服务器模型在不同客户端上表现不一，可能偏向某些客户端而忽视其他，加剧了公平性挑战。本文重新审视客户端性能分布不一致问题，引入对抗性多臂老虎机方法，对目标函数施加显式性能差异约束。提出一种基于多臂老虎机的新型联邦学习分配算法（FedMABA），以缓解具有不同数据分布的客户端间的性能不公平。在多种非独立同分布（Non-I.I.D.）场景下的大量实验表明，FedMABA在提升公平性方面表现优异。
原文摘要 · Abstract (English)
The increasing concern for data privacy has driven the rapid development of federated learning (FL), a privacy-preserving collaborative paradigm. However, the statistical heterogeneity among clients in FL results in inconsistent performance of the server model across various clients. Server model may show favoritism towards certain clients while performing poorly for others, heightening the challenge of fairness. In this paper, we reconsider the inconsistency in client performance distribution and introduce the concept of adversarial multi-armed bandit to optimize the proposed objective with explicit constraints on performance disparities. Practically, we propose a novel multi-armed bandit-based allocation FL algorithm (FedMABA) to mitigate performance unfairness among diverse clients with different data distributions. Extensive experiments, in different Non-I.I.D. scenarios, demonstrate the exceptional performance of FedMABA in enhancing fairness.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。