FedRGL: Robust Federated Graph Learning for Label Noise

TL;DR

FedRGL is a novel federated graph learning approach that effectively handles label noise by combining dual-perspective filtering, contrastive learning, and adaptive aggregation, leading to superior performance on real-world datasets.

Contribution

This paper introduces FedRGL, a robust federated graph learning method that addresses label noise using dual-perspective filtering, contrastive learning, and entropy-based adaptive aggregation.

Findings

Outperforms 12 baseline methods across multiple datasets.

Effectively handles various noise rates and types.

Improves model robustness and generalization in federated settings.

Abstract

Federated Graph Learning (FGL) is a distributed machine learning paradigm based on graph neural networks, enabling secure and collaborative modeling of local graph data among clients. However, label noise can degrade the global model's generalization performance. Existing federated label noise learning methods, primarily focused on computer vision, often yield suboptimal results when applied to FGL. To address this, we propose a robust federated graph learning method with label noise, termed FedRGL. FedRGL introduces dual-perspective consistency noise node filtering, leveraging both the global model and subgraph structure under class-aware dynamic thresholds. To enhance client-side training, we incorporate graph contrastive learning, which improves encoder robustness and assigns high-confidence pseudo-labels to noisy nodes. Additionally, we measure model quality via predictive entropy of unlabeled nodes, enabling adaptive robust aggregation of the global model. Comparative experiments on multiple real-world graph datasets show that FedRGL outperforms 12 baseline methods across various noise rates, types, and numbers of clients.