提出公平且可扩展的联邦遗忘方法，解决真实场景下数据不均衡问题。
Federated Unlearning in the Wild: Rethinking Fairness and Data Discrepancy
- 设计跨客户端约束机制，按需执行遗忘，避免全体重训
- 在真实数据异构下性能优于现有方法，遗忘后模型准确率下降<5%
- 适合需要合规删除数据的医疗、金融等隐私敏感领域
机器遗忘对于实现数据删除权（如“被遗忘权”）至关重要。作为去中心化范式，联邦学习（FL）同样需要遗忘能力，但实际应用面临两大挑战：一是联邦遗忘（FU）中的公平性常被忽视，精确方法强制所有客户端重训，即使未参与训练；近似方法通过梯度上升或知识蒸馏粗粒度干预，可能不公平地损害仅保留数据客户端的性能。二是现有评估多依赖合成数据假设（独立同分布/非独立同分布），忽略真实世界异构性，导致方法真实影响被掩盖。我们首次在真实数据异构与公平性条件下全面评估现有FU方法。随后提出新型公平感知的联邦遗忘方法——联邦跨客户端约束遗忘（FedCCCU），显式应对上述挑战。实验表明，在真实设置中，现有方法表现不佳，而本方法始终显著优于基线。
原文摘要 · Abstract (English)
Machine unlearning is critical for enforcing data deletion rights like the "right to be forgotten." As a decentralized paradigm, Federated Learning (FL) also requires unlearning, but realistic implementations face two major challenges. First, fairness in Federated Unlearning (FU) is often overlooked. Exact unlearning methods typically force all clients into costly retraining, even those uninvolved. Approximate approaches, using gradient ascent or distillation, make coarse interventions that can unfairly degrade performance for clients with only retained data. Second, most FU evaluations rely on synthetic data assumptions (IID/non-IID) that ignore real-world heterogeneity. These unrealistic benchmarks obscure the true impact of unlearning and limit the applicability of current methods. We first conduct a comprehensive benchmark of existing FU methods under realistic data heterogeneity and fairness conditions. We then propose a novel, fairness-aware FU approach, Federated Cross-Client-Constrains Unlearning (FedCCCU), to explicitly address both challenges. FedCCCU offers a practical and scalable solution for real-world FU. Experimental results show that existing methods perform poorly in realistic settings, while our approach consistently outperforms them.
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。