AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Federated Graph Learning via Structure-Aware Fusion Using a Kalman Framework with Learnable Dynamics

Bisheng Tang, Xiaojun Chen

ICML 2026 · Graph learning › Graph self-supervised learning · approach: Heterophily and imbalance on graphs

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use Fed-Kalter instead of FedStar across 1 to 4 domain settingsfederated graph classification · 1-4 domain settingsaverage accuracy: 5.56%, 2.57%, 1.00%, and 1.46% over FedStarImproved
use Fed-Kalter instead of FedVN across 1 to 4 domain settingsfederated graph classification · 1-4 domain settingsaverage accuracy: 3.95%, 2.58%, 0.82%, and 0.68%Improved
use Fed-Kalter instead of FedAvgfederated graph classification · One-domain averageaccuracy: 70.23 -> 74.65Improved
use Fed-Kalter instead of FedAvgfederated graph classification · Four-domain CHEM-BIO-SN-CVaccuracy: 66.75 -> 73.69Improved
use federated training (FedAvg) instead of local-only traininggraph classification · Four-domain CHEM-BIO-SN-CVaccuracy: 71.37 -> 66.75Hurt
use Fed-Kalter instead of local-only traininggraph classification · Four-domain CHEM-BIO-SN-CVaccuracy: 71.37 -> 73.69Improved
use Fed-Kalter instead of FedSage on computer vision graphsfederated graph classification · One-domain CVaccuracy: 85.78 -> 85.38No clear effect
use Fed-Kalter instead of FedStar on social networksfederated graph classification · One-domain SNaccuracy: 63.68 -> 68.14Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy