提出无需控制变量的通信压缩框架，解决联邦学习中的隐私与状态难题。
Communication Compression for Distributed Learning without Control Variates
- 利用历史聚合更新实现高精度压缩，避免依赖客户端控制变量。
- 理论证明在非凸场景下优于传统有偏压缩方法，收敛性更优。
- 适合隐私敏感、资源受限的分布式学习场景，如联邦学习。
分布式学习算法（如联邦学习）需通过通信压缩降低客户端上传开销。现有压缩方法常引入偏差，需误差反馈以保证收敛和理论保障，但误差反馈依赖客户端特定的控制变量，既违反隐私保护原则，又要求客户端维持状态。本文提出压缩聚合反馈（CAFe）框架，通过利用历史聚合更新实现高度可压缩的客户端更新，且无需控制变量。以分布式梯度下降（DGD）为例，理论证明在非凸场景且梯度差异有界的条件下，CAFe优于带有有偏压缩的分布式压缩梯度下降（DCGD）。实验结果验证了CAFe在多种压缩方案中表现更优。
原文摘要 · Abstract (English)
Distributed learning algorithms, such as the ones employed in Federated Learning (FL), require communication compression to reduce the cost of client uploads. The compression methods used in practice are often biased, making error feedback necessary both to achieve convergence under aggressive compression and to provide theoretical convergence guarantees. However, error feedback requires client-specific control variates, creating two key challenges: it violates privacy-preserving principles and demands stateful clients. In this paper, we propose Compressed Aggregate Feedback (CAFe), a novel distributed learning framework that allows highly compressible client updates by exploiting past aggregated updates, and does not require control variates. We consider Distributed Gradient Descent (DGD) as a representative algorithm and analytically prove CAFe's superiority to Distributed Compressed Gradient Descent (DCGD) with biased compression in the non-convex regime with bounded gradient dissimilarity. Experimental results confirm that CAFe outperforms existing distributed learning compression schemes.
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。