Research map / Federated learning and privacy / Personalized and heterogeneous FL
FedSoft: Soft Clustered Federated Learning with Proximal Local Updating
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 FedSoft instead of IFCA | cluster and personalized modeling on EMNIST Letters · 10:90 and linear partitions | accuracy (%): 70.3 vs 58.5; 74.6 vs 61.3; 90.9 vs 65.2; 71.8 vs 55.4; 74.1 vs 57.2; 86.5 vs 62.9 | Helped |
| use FedSoft instead of FedEM | cluster and personalized modeling on EMNIST Letters · 10:90 and linear partitions | accuracy (%): 70.3 vs 67.4; 74.6 vs 69.8; 90.9 vs 63.6; 71.8 vs 65.9; 74.1 vs 69.0; 86.5 vs 62.4 | Helped |
| add proximal regularization with λ=0.1 rather than setting λ=0 (ablation) | FedSoft · EMNIST Letters, linear partition | accuracy (%): 48.4 -> 71.8; 47.7 -> 74.1; 72.7 -> 86.5 | Helped |
| increase proximal regularization weight from λ=0.1 to λ=1 (ablation) | FedSoft · EMNIST Letters, linear partition | accuracy (%): 71.8 -> 55.1; 74.1 -> 58.7; 86.5 -> 63.6 | Hurt |
| increase distribution divergence by setting σ₀ from 1 to 10 (ablation) | FedSoft · Synthetic, random partition | MSE: 5.9 -> 42.2; 4.2 -> 27.0; 0.6 -> 4.96 | Hurt |
| increase distribution divergence by setting σ₀ from 10 to 50 and 100 (ablation) | FedSoft · Synthetic, random partition | MSE: 42.2 -> 225.3; 60.6 -> 89.9; 42.7 -> 117.6; 27.0 -> 89.9; 4.96 -> 18.8; 42.2 -> 782.4; 60.6 -> 454.4; 42.7 -> 432.8; 27. | Hurt |
Closest papers by meaning
- FLUX: Efficient Descriptor-Driven Clustered Federated Learning under Arbitrary Distribution Shifts (NeurIPS 2025 · similarity 0.88)
- FedImpro: Measuring and Improving Client Update in Federated Learning (ICLR 2024 · similarity 0.88)
- FedAS: Bridging Inconsistency in Personalized Federated Learning (CVPR 2024 · similarity 0.88)
- Clustered Federated Learning via Gradient-based Partitioning (ICML 2024 · similarity 0.88)
- Enhancing Clustered Federated Learning: Integration of Strategies and Improved Methodologies (ICLR 2025 · similarity 0.88)
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · similarity 0.88)
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.