GraphFedMIG: Tackling Class Imbalance in Federated Graph Learning via Mutual Information-Guided Generation

TL;DR

GraphFedMIG introduces a federated graph learning framework that uses mutual information-guided generative data augmentation to address class imbalance and improve model performance on decentralized, non-IID graph data.

Contribution

It proposes a novel federated generative augmentation method with mutual information guidance to enhance minority class learning in federated graph neural networks.

Findings

Outperforms baseline methods on four real-world datasets.

Effectively mitigates class imbalance in federated graph learning.

Improves minority class representation and model accuracy.

Abstract

Federated graph learning (FGL) enables multiple clients to collaboratively train powerful graph neural networks without sharing their private, decentralized graph data. Inherited from generic federated learning, FGL is critically challenged by statistical heterogeneity, where non-IID data distributions across clients can severely impair model performance. A particularly destructive form of this is class imbalance, which causes the global model to become biased towards majority classes and fail at identifying rare but critical events. This issue is exacerbated in FGL, as nodes from a minority class are often surrounded by biased neighborhood information, hindering the learning of expressive embeddings. To grapple with this challenge, we propose GraphFedMIG, a novel FGL framework that reframes the problem as a federated generative data augmentation task. GraphFedMIG employs a hierarchical generative adversarial network where each client trains a local generator to synthesize high-fidelity feature representations. To provide tailored supervision, clients are grouped into clusters, each sharing a dedicated discriminator. Crucially, the framework designs a mutual information-guided mechanism to steer the evolution of these client generators. By calculating each client's unique informational value, this mechanism corrects the local generator parameters, ensuring that subsequent rounds of mutual information-guided generation are focused on producing high-value, minority-class features. We conduct extensive experiments on four real-world datasets, and the results demonstrate the superiority of the proposed GraphFedMIG compared with other baselines.