Fed-GAME: Personalized Federated Learning with Graph Attention Mixture-of-Experts For Time-Series Forecasting

TL;DR

Fed-GAME introduces a dynamic graph-based personalized federated learning framework for time-series forecasting, effectively handling client heterogeneity and improving accuracy over existing methods.

Contribution

The paper proposes Fed-GAME, a novel personalized federated learning framework using a learnable dynamic graph and a Graph Attention Mixture-of-Experts aggregator for improved time-series prediction.

Findings

Outperforms state-of-the-art personalized FL methods on electric vehicle datasets.

Effectively models client heterogeneity with a dynamic implicit graph.

Enhances time-series forecasting accuracy in federated settings.

Abstract

Federated learning (FL) on graphs shows promise for distributed time-series forecasting. Yet, existing methods rely on static topologies and struggle with client heterogeneity. We propose Fed-GAME, a framework that models personalized aggregation as message passing over a learnable dynamic implicit graph. The core is a decoupled parameter difference-based update protocol, where clients transmit parameter differences between their fine-tuned private model and a shared global model. On the server, these differences are decomposed into two streams: (1) averaged difference used to updating the global model for consensus (2) the selective difference fed into a novel Graph Attention Mixture-of-Experts (GAME) aggregator for fine-grained personalization. In this aggregator, shared experts provide scoring signals while personalized gates adaptively weight selective updates to support personalized aggregation. Experiments on two real-world electric vehicle charging datasets demonstrate that Fed-GAME outperforms state-of-the-art personalized FL baselines.