Research map / Federated learning and privacy / Communication-efficient FL
Federated Learning with Matched Averaging
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use layer-wise matched averaging instead of FedAvg | VGG-9 · CIFAR-10 (heterogeneous) | accuracy (%): 86.29 -> 87.53 | Helped |
| Use layer-wise matched averaging instead of FedProx | VGG-9 · CIFAR-10 (heterogeneous) | accuracy (%): 85.32 -> 87.53 | Helped |
| Use layer-wise matched averaging instead of FedAvg | 1-layer LSTM · Shakespeare | accuracy (%): 46.63 -> 49.07 | Helped |
| Use layer-wise matched averaging instead of FedProx | 1-layer LSTM · Shakespeare | accuracy (%): 45.83 -> 49.07 | Helped |
| Use recursive layer-by-layer matching rather than the proposed layer-wise freeze-and-retrain scheme (ablation) | deep CNN matching · CIFAR-10, VGG-9 | accuracy | Hurt |
| Extend PFNM to CNNs using matched averaging (ablation) | LeNet · MNIST | accuracy | Helped |
| Extend PFNM to CNNs using matched averaging (ablation) | VGG-9 · CIFAR-10 | accuracy | Hurt |
| Increase local training epochs (ablation) | FedMA · CIFAR-10 (heterogeneous), VGG-9 | accuracy | Helped |
| Increase local training epochs (ablation) | FedAvg · CIFAR-10 (heterogeneous), VGG-9 | accuracy | Hurt |
| Increase local training epochs (ablation) | FedProx · CIFAR-10 (heterogeneous), VGG-9 | accuracy | Hurt |
| Use FedMA rather than centralized training on the skewed dataset | VGG-9 · Skewed CIFAR-10 | accuracy | Helped |
| Add clients with additional heterogeneous data (ablation) | FedMA · CIFAR-10 | accuracy | Helped |
Closest papers by meaning
- FedQS: Optimizing Gradient and Model Aggregation for Semi-Asynchronous Federated Learning (NeurIPS 2025 · similarity 0.86)
- Rethinking the Flat Minima Searching in Federated Learning (ICML 2024 · similarity 0.85)
- FedASMU: Efficient Asynchronous Federated Learning with Dynamic Staleness-Aware Model Update (AAAI 2024 · similarity 0.85)
- Progressive Distribution Matching for Federated Semi-Supervised Learning (AAAI 2025 · similarity 0.85)
- Beyond Local Sharpness: Communication-Efficient Global Sharpness-aware Minimization for Federated Learning (CVPR 2025 · similarity 0.85)
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · similarity 0.85)
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.