在保护隐私的前提下，提升个性化联邦学习的收敛性与公平性
Convergence-Privacy-Fairness Trade-Off in Personalized Federated Learning
- 基于差分隐私改进经典Ditto算法，实现隐私-收敛-公平三者权衡
- 理论推导最优全局聚合次数，在给定隐私预算下提升性能
- 实验显示比现有先进方法公平性提升超32.71%，准确率高9.66%
个性化联邦学习（PFL）如著名算法Ditto，通过联邦学习（FL）指导个性化学习（PL），平衡个性化与泛化。然而，客户端为保护隐私对本地模型加噪，会影响PL的收敛性和性能公平性。本文提出在差分隐私（DP）保护下的PFL方法DP-Ditto，分析其隐私保障、模型收敛与性能公平之间的权衡关系。推导了在DP-Ditto下个性化模型的收敛上界，给出给定隐私预算下的最优全局聚合次数。进一步分析个性化模型的性能公平性，证明可联合优化收敛性与公平性。实验验证分析有效性，表明DP-Ditto在公平性上超越当前最先进的PFL模型（如FedAMP、pFedMe、APPLE、FedALA）超过32.71%，准确率提升9.66%。
原文摘要 · Abstract (English)
Personalized federated learning (PFL), e.g., the renowned Ditto, strikes a balance between personalization and generalization by conducting federated learning (FL) to guide personalized learning (PL). While FL is unaffected by personalized model training, in Ditto, PL depends on the outcome of the FL. However, the clients' concern about their privacy and consequent perturbation of their local models can affect the convergence and (performance) fairness of PL. This paper presents PFL, called DP-Ditto, which is a non-trivial extension of Ditto under the protection of differential privacy (DP), and analyzes the trade-off among its privacy guarantee, model convergence, and performance distribution fairness. We also analyze the convergence upper bound of the personalized models under DP-Ditto and derive the optimal number of global aggregations given a privacy budget. Further, we analyze the performance fairness of the personalized models, and reveal the feasibility of optimizing DP-Ditto jointly for convergence and fairness. Experiments validate our analysis and demonstrate that DP-Ditto can surpass the DP-perturbed versions of the state-of-the-art PFL models, such as FedAMP, pFedMe, APPLE, and FedALA, by over 32.71% in fairness and 9.66% in accuracy.
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。