让个性化联邦学习更公平，避免有害同伴拖后腿
SP-CACW: Convergence-Aware Client Weighting for Selfish Personalized Learning
- 根据收敛误差上界动态分配同伴权重，自动剔除有害参与方
- 在多个数据集上表现优于或媲美现有顶尖个性化方法
- 适合数据分布差异大、需自我保护的个性化学习场景
协作学习的可持续性取决于每个参与方都能获益。标准联邦学习优化全局平均目标，当客户端数据分布与整体差异较大时性能下降。本文研究自私个性化：如何让指定目标客户端利用同伴梯度最小化自身风险，同时避免负迁移。提出SP-CACW框架，通过最小化目标客户端收敛误差上界来选择聚合权重，显式权衡同伴偏差与随机方差，可将有害同伴权重设为零。在光滑性和有界方差假设下提供收敛保证，并在MNIST、CIFAR-100和LEAF Shakespeare数据集上评估，表现优于或媲美强个性化与聚类基线。
原文摘要 · Abstract (English)
Collaborative learning is sustainable only when it benefits each participant. Standard federated learning optimizes a global average objective, which can under perform for clients whose data distributions differ substantially from the population. We study selfish personalization: how a designated target client can use peer gradients to minimize its own risk while avoiding negative transfer. We propose SP-CACW, a convergence-aware client-weighting framework that selects aggregation weights by minimizing an upper bound on the target client's convergence error. The resulting rule explicitly trades off peer bias against stochastic variance and can assign zero weight to harmful peers. We provide convergence guarantees under smoothness and bounded-variance assumptions and evaluate the method on MNIST, CIFAR-100, and LEAF Shakespeare, where it is competitive with or improves over strong personalized and clustering baselines.
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。