Saved in:
Bibliographic Details
Main Authors: Li, Yi, Liu, Han, Fan, Mingfeng, Chen, Guo, Li, Chaojie, Sikdar, Biplab
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.01363
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910037608235008
author Li, Yi
Liu, Han
Fan, Mingfeng
Chen, Guo
Li, Chaojie
Sikdar, Biplab
author_facet Li, Yi
Liu, Han
Fan, Mingfeng
Chen, Guo
Li, Chaojie
Sikdar, Biplab
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2603_01363
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Fed-GAME: Personalized Federated Learning with Graph Attention Mixture-of-Experts For Time-Series Forecasting
Li, Yi
Liu, Han
Fan, Mingfeng
Chen, Guo
Li, Chaojie
Sikdar, Biplab
Machine Learning
Distributed, Parallel, and Cluster Computing
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.
title Fed-GAME: Personalized Federated Learning with Graph Attention Mixture-of-Experts For Time-Series Forecasting
topic Machine Learning
Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2603.01363