AI Paper Atlas

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

FedBR: Improving Federated Learning on Heterogeneous Data via Local Learning Bias Reduction

Yongxin Guo, Xiaoying Tang, Tao Lin

ICML 2023 · Federated learning and privacy › Communication-efficient FL · approach: Local drift correction · 8 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 classifier-output balancing to FedAvg (ablation)FedAvg · RotatedMNISTaccuracy (%): 82.47 -> 84.40Improved
Add classifier-output balancing to FedAvg (ablation)FedAvg · CIFAR10accuracy (%): 58.99 -> 64.32Improved
Add classifier-output balancing to FedAvg (ablation)FedAvg · CIFAR100accuracy (%): 46.37 -> 47.22Improved
Add adversarial feature debiasing to FedAvg (ablation)FedAvg · RotatedMNISTaccuracy (%): 82.47 -> 86.25Improved
Add adversarial feature debiasing to FedAvg (ablation)FedAvg · CIFAR10accuracy (%): 58.99 -> 63.44Improved
Add adversarial feature debiasing to FedAvg (ablation)FedAvg · CIFAR100accuracy (%): 46.37 -> 47.78Improved
Add the max step to adversarial feature debiasing (ablation)Component 2 · RotatedMNISTaccuracy (%): 81.24 -> 86.25Improved
Add the max step to adversarial feature debiasing (ablation)Component 2 · CIFAR10accuracy (%): 58.84 -> 63.44Improved
Add the max step to adversarial feature debiasing (ablation)Component 2 · CIFAR100accuracy (%): 43.50 -> 47.78Improved
Combine classifier-output balancing with adversarial feature debiasing (ablation)FedBR · RotatedMNISTaccuracy (%): Component 1: 84.40; Component 2: 86.25; FedBR: 86.58Improved
Combine classifier-output balancing with adversarial feature debiasing (ablation)FedBR · CIFAR10accuracy (%): Component 1: 64.32; Component 2: 63.44; FedBR: 64.65Improved
Combine classifier-output balancing with adversarial feature debiasing (ablation)FedBR · CIFAR100accuracy (%): Component 1: 47.22; Component 2: 47.78; FedBR: 47.75No clear effect

Later papers that compared with this method

1 later accepted paper reports FedBR as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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