AI Paper Atlas

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

Federated Learning with Manifold Regularization and Normalized Update Reaggregation

Xuming An, Li Shen, Han Hu, Yong Luo

NeurIPS 2023 · Federated learning and privacy › Communication-efficient FL · approach: Local drift correction

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 normalized update aggregation (FedMRUR-N) (ablation)MoFedSAM · CIFAR-100 Dir(0.3)test accuracy: 52.81 -> 54.27Improved
add hyperbolic graph fusion manifold regularization (FedMRUR-H) (ablation)MoFedSAM · CIFAR-100 Dir(0.3)test accuracy: 52.81 -> 53.57Improved
combine normalization and hyperbolic regularization (full FedMRUR) (ablation)MoFedSAM · CIFAR-100 Dir(0.3)test accuracy: 52.81 -> 55.49Improved
FedMRUR vs second-best on CIFAR-100 Dir(0.3)FedMRUR · CIFAR-100 Dir(0.3)test accuracy: 52.81 -> 55.49Improved
FedMRUR vs MoFedSAM on TinyImageNet Path(40)FedMRUR · TinyImageNet Path(40)test accuracy: 41.58 -> 45.71Improved
FedMRUR vs MoFedSAM on CIFAR-10 Dir(0.3)FedMRUR · CIFAR-10 Dir(0.3)test accuracy: 84.03 -> 84.53Improved
FedMRUR vs FedCM on rounds to reach 38% on CIFAR-100 Dir(0.3)FedMRUR · CIFAR-100 Dir(0.3)communication rounds: 126 -> 179Hurt
FedMRUR vs FedAvg on rounds to 38% on CIFAR-100 Dir(0.3)FedMRUR · CIFAR-100 Dir(0.3)communication rounds: 494 -> 179Improved
FedMRUR vs MoFedSAM, rounds to 45% on CIFAR-100 Dir(0.3)FedMRUR · CIFAR-100 Dir(0.3)communication rounds: 278 -> 263Improved
FedMRUR vs MoFedSAM, rounds to 80% on CIFAR-10 Dir(0.3)FedMRUR · CIFAR-10 Dir(0.3)communication rounds: 450 -> 449No clear effect
raise client participation ratio from 0.02 to 0.2 (ablation)FedMRUR · CIFAR-100 Dir(0.3)convergence speed / test accuracyMixed
vary gamma in [0.005, 0.5] (ablation)FedMRUR · CIFAR-100test accuracy / training lossNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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