AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

MSCFL: Model Structure-Aware Clustered Federated Learning for System Heterogeneity and Data Drift

Yang Xu, Xiaowei Wu, Zifeng Xu, Cheng Zhang, Ju Ren, Yaoxue Zhang

AAAI 2026 · Federated learning and privacy › Communication-efficient FL · approach: Efficient local training in FL · 0 citations (OpenAlex)

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
add cluster migration strategy (Hamming-distance drift detection) (ablation)MSCFL · CIFAR-10 under feature/label driftaccuracyImproved
add cluster migration strategy (ablation)MSCFL · CIFAR100 clustering timetime (s): 3.58 -> 10.18Hurt
add cluster migration strategy (ablation)MSCFL · SVHN clustering timetime (s): 0.72 -> 1.08Hurt
use model-structure similarity (MSCFL) instead of FL+HC on pruned modelsMSCFL · CIFAR10 Feature Non-IIDaccuracy (%): 54.29 -> 64.38Improved
use MSCFL vs IFCA under label Non-IIDMSCFL · FMNIST Label Non-IIDaccuracy (%): 99.45 -> 99.37No clear effect
use MSCFL vs IFCA under label Non-IIDMSCFL · CIFAR10 Label Non-IIDaccuracy (%): 83.10 -> 91.84Improved
use structure-masked similarity vs plain shared-part or NXOR similarity when combining CFL with MP (ablation)MSCFL · CIFAR-10 clustering accuracyclustering accuracyImproved
MSCFL clustering time vs IFCAMSCFL · CIFAR100 clustering timetime (s): 82.22 -> 10.18Improved

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