Research map / Federated learning and privacy / Communication-efficient FL
DFedADMM: Dual Constraint Controlled Model Inconsistency for Decentralize Federated Learning
Qinglun Li, Li Shen, Guanghao Li, Quanjun Yin, Dacheng Tao
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add SAM local optimizer to DFedADMM (ablation) | DFedADMM · CIFAR-10 (Dir 0.3) | top-1 accuracy (%): 80.50 -> 82.00 | Improved |
| add SAM local optimizer to DFedADMM (ablation) | DFedADMM · CIFAR-100 (Dir 0.3) | top-1 accuracy (%): 58.83 -> 59.37 | Improved |
| add SAM local optimizer to DFedADMM (ablation) | DFedADMM · CIFAR-10 (IID) | top-1 accuracy (%): 80.05 -> 83.92 | Improved |
| add SAM local optimizer to DFedADMM (ablation) | DFedADMM · MNIST (Dir 0.3) | top-1 accuracy (%): 98.07 -> 98.44 | Improved |
| replace DFedSAM with dual-constrained DFedADMM-SAM | DFedSAM · CIFAR-10 (Dir 0.3) | top-1 accuracy (%): 80.10 -> 82.00 | Improved |
| replace DFedSAM with dual-constrained DFedADMM-SAM | DFedSAM · CIFAR-100 (Dir 0.3) | top-1 accuracy (%): 59.19 -> 59.37 | Improved |
| replace centralized FedSAM with decentralized DFedADMM-SAM | FedSAM · CIFAR-10 (Dir 0.3) | top-1 accuracy (%): 80.56 -> 82.00 | Improved |
| make the topology better connected (Ring to Full) | DFedADMM-SAM · MNIST (topology) | top-1 accuracy (%): 97.08 -> 98.50 | Improved |
| make the topology better connected (Ring to Full) | DFedADMM · MNIST (topology) | top-1 accuracy (%): 96.41 -> 98.26 | Improved |
| increase data heterogeneity from IID to Dir 0.3 | DFedADMM-SAM · CIFAR-10 | top-1 accuracy (%): 1.92% | Hurt |
| increase local iterations K to 10 (from 1, 2, 5) (ablation) | DFedADMM-SAM · CIFAR-10 | convergence speed | Improved |
| set perturbation radius rho=0.1 versus 0.01 to 0.3 (ablation) | DFedADMM-SAM · CIFAR-10 | accuracy | Mixed |
Closest papers by meaning
- Decentralized Sporadic Federated Learning: A Unified Algorithmic Framework with Convergence Guarantees (ICLR 2025 · similarity 0.91)
- SemiDFL: A Semi-Supervised Paradigm for Decentralized Federated Learning (AAAI 2025 · similarity 0.89)
- NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel (ICML 2025 · similarity 0.89)
- Overcoming Data and Model heterogeneities in Decentralized Federated Learning via Synthetic Anchors (ICML 2024 · similarity 0.88)
- Decentralized Directed Collaboration for Personalized Federated Learning (CVPR 2024 · similarity 0.88)
- FedSMU: Communication-Efficient and Generalization-Enhanced Federated Learning through Symbolic Model Updates (ICML 2025 · 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.
