arXiv：cs.AI· Dario Fenoglio, Daniil Kirilenko, Martin Gjoreski, Marc Langheinrich·· 23 小时前
FedCRF：用条件随机场结构化概率推断实现联邦学习聚合
Federated Learning with Energy-Based Structured Probabilistic Inference
AI 导读
FedCRF 将联邦学习中的服务器聚合建模为基于条件随机场的结构化概率推断，把客户端表示为图中的节点，用一元势函数刻画客户端自身属性、成对势函数刻画客户端间相似性。推断出的客户端可靠性与亲和度可统一支撑全局聚合、拜占庭鲁棒、聚类联邦与个性化联邦，单一超参数配置下在 CIFAR-10、FEMNIST 和 Fed-ISIC2019 上媲美或超越专用方法。
正文
Abstract:Federated Learning (FL) enables multiple clients to collaboratively train machine learning models without sharing their local data. However, heterogeneity in client data and the presence of unreliable or adversarial updates make effective server aggregation challenging, motivating methods that adapt client contributions according to their reliability and relationships. In this work, we formulate server aggregation in FL as structured probabilistic inference using conditional random fields. We represent clients as nodes in a graph and define unary and pairwise energy potentials. Unary potentials capture individual client properties, while pairwise potentials encode similarities between clients. This formulation treats individual client evidence and pairwise relationships as coupled evidence in a joint inference problem, allowing client reliabilities to be inferred collectively and the resulting estimates to guide aggregation. Where most FL methods target a single challenge, FedCRF's inferred reliabilities and affinities provide one common representation for global aggregation, Byzantine robustness, clustered FL, and personalized FL. With a single set of hyperparameters, it is competitive with or outperforms specialized methods on CIFAR-10, FEMNIST, and Fed-ISIC2019 across diverse forms of client heterogeneity and unreliability.
| Comments: | Revised version (23 pages) |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.30161 [cs.LG] |
| (or arXiv:2606.30161v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2606.30161 arXiv-issued DOI via DataCite |
Submission history
From: Dario Fenoglio [view email]
[v1]
Mon, 29 Jun 2026 11:38:46 UTC (165 KB)
[v2]
Fri, 9 Oct 2026 19:57:18 UTC (194 KB)
来源：arXiv：cs.AI · arxiv.org