Abstract:Collaborative machine learning (CML) enables multiple clients to train a global model jointly in a data-distributed setting. To address data privacy and communication efficiency, one-shot CML has been increasingly adopted, where clients communicate with the server only once by sharing synthetic or processed proxy data. This single-round communication, however, eliminates the possibility of iterative correction at the server, making the learning process particularly vulnerable to client unreliability. In this setting, unreliable clients, whether malicious or non-malicious, may provide biased proxy data that favors certain groups, thereby degrading the fairness of the global model and harming minority or unprivileged groups. In this work, we propose a server-side defense framework based on a bilevel optimization formulation. The proposed approach learns client-level weights to mitigate the influence of biased client proxy data while enforcing fairness constraints by using a very small trusted root dataset available at the server. Experimental results on benchmark datasets show that our method improves fairness with little accuracy loss under biased proxy data contributions from unreliable clients. Moreover, the proposed approach remains effective even when unreliable clients make up a majority of the system, consistently outperforming other existing methods.
| Comments: | Accepted at the 2nd International Conference on Federated Learning and Intelligent Computing Systems (FLICS 2026) |
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.08616 [cs.LG] |
| (or arXiv:2605.08616v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.08616 arXiv-issued DOI via DataCite (pending registration) |
From: Chia-Yuan Wu [view email]
[v1]
Sat, 9 May 2026 02:19:31 UTC (1,552 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。