AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID Data

Mi Luo, Fei Chen, Dapeng Hu, Yifan Zhang, Jian Liang, Jiashi Feng

NeurIPS 2021 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Non-IID federated learning · 39 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 post-hoc classifier calibration with virtual representations (CCVR)FedAvg · CINIC-10Accuracy@1: 60.20 -> 69.99Improved
add CCVRFedAvg · CIFAR-10 alpha=0.1Accuracy@1: 58.55 -> 62.68Improved
add CCVRMOON · CIFAR-10 alpha=0.05Accuracy@1: 49.91 -> 55.60Improved
add CCVRFedAvg · CIFAR-100Accuracy@1: 66.25 -> 66.60Improved
add classifier weight L2-normalization (clsnorm) during training (ablation)FedAvg · CIFAR-10 alpha=0.05Accuracy@1: 68.62 -> 69.65 (alpha=0.5); 52.33 -> 51.74 (alpha=0.05)Mixed
add classifier proximal regularization (clsprox) during training (ablation)FedAvg · CIFAR-10 alpha=0.1Accuracy@1: 58.55 -> 59.04No clear effect
post-calibrate classifier with real IID data (oracle) (ablation)FedAvg · CIFAR-10 alpha=0.1Accuracy@1: 58.55 -> 64.70Improved
increase virtual features per class from 50 to 2000 (ablation)CCVR on FedAvg · CIFAR-10 alpha=0.1Accuracy@1: 61.07 -> 62.68Improved
increase virtual features per class (50 to 1000) (ablation)CCVR on FedAvg · CIFAR-10 alpha=0.5Accuracy@1: 70.99 -> 70.71Mixed
increase number of clients 10 to 100 (ablation)CCVR on FedAvg · CIFAR-10 alpha=0.1Accuracy@1: 58.55 -> 62.68 (N=10); 55.42 -> 59.19 (N=100)No clear effect

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