Research map / Federated learning and privacy / Personalized and heterogeneous FL
Federated Learning with Label Distribution Skew via Logits Calibration
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 changed | Where | Measured | Effect |
|---|---|---|---|
| calibrate local logits with class-count-based pairwise margins | FedAvg-style federated learning · CIFAR10, D(0.05) | test accuracy: 44.39±1.63 -> 54.55±1.70 | Improved |
| calibrate local logits with class-count-based pairwise margins | FedAvg-style federated learning · CIFAR10, Q(2) | test accuracy: 55.91±1.44 -> 61.87±0.94 | Improved |
| calibrate local logits with class-count-based pairwise margins | federated learning · FEMNIST | test accuracy: 87.95±0.92 -> 92.78±0.58 | Improved |
| add FedProx regularization to FedLC (ablation) | FedLC · CIFAR10, α=5 | test accuracy: 69.89 -> 71.39 | Improved |
| add Scaffold to FedLC (ablation) | FedLC · SVHN, α=5 | test accuracy: 88.19 -> 89.34 | Improved |
| increase local epochs from E=1 to E=5 (ablation) | FedLC · CIFAR10, β=0.5 | test accuracy: 70.41 -> 80.25 | Improved |
| increase number of clients from m=10 to m=100 (ablation) | FedLC · reported test accuracy setting | test accuracy: 70.13 -> 43.81 | Hurt |
Closest papers by meaning
- Exploring Vacant Classes in Label-Skewed Federated Learning (AAAI 2025 · similarity 0.86)
- Exploiting Label Skews in Federated Learning with Model Concatenation (AAAI 2024 · similarity 0.84)
- Progressive Distribution Matching for Federated Semi-Supervised Learning (AAAI 2025 · similarity 0.83)
- FedClean: A General Robust Label Noise Correction for Federated Learning (ICML 2025 · similarity 0.83)
- FedWon: Triumphing Multi-domain Federated Learning Without Normalization (ICLR 2024 · similarity 0.82)
- A Simple Data Augmentation for Feature Distribution Skewed Federated Learning (CVPR 2025 · similarity 0.82)
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.
