Research map / Federated learning and privacy / Communication-efficient FL
DecFus: Decentralized Layer-wise Fusion with Dynamic Exploration and Exploitation
Li Yang, Jialong Sun, Chuhai Cai, Xinyang Liu, Yichen Li, Bowen Peng, Jialong Li, Bo Liu
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 dynamic cutoff (exchange then averaging) instead of reversed cutoff (ablation) | DecFus · CIFAR-100 (alpha=0.1), ResNet-50 | accuracy: 26.28 -> 35.76 | Improved |
| add dynamic cutoff instead of fixed high cutoff (ablation) | DecFus · SVHN (alpha=0.1), ResNet-50 | accuracy: 52.68 -> 57.83 | Improved |
| add dynamic cutoff instead of fixed low cutoff (ablation) | DecFus · CIFAR-100 (alpha=0.1), ResNet-50 | accuracy: 30.50 -> 35.76 | Improved |
| use cosine similarity instead of L2 distance (ablation) | DecFus · CIFAR-100 (alpha=0.1), ResNet-50 | accuracy: 33.05 -> 35.76 | Improved |
| use cosine similarity instead of L2 distance (ablation) | DecFus · CIFAR-10 (IID), ResNet-50 | accuracy: 78.75 -> 78.24 | No clear effect |
| use cosine similarity instead of KL divergence (ablation) | DecFus · SVHN (alpha=0.1), ResNet-50 | accuracy: 53.58 -> 57.83 | Improved |
| use similarity-based grouping instead of random grouping (ablation) | DecFus · CIFAR-10 (alpha=0.1), VGG16 | accuracy: 48.36 -> 54.14 | Improved |
| use similarity-based grouping instead of inverse grouping (ablation) | DecFus · CIFAR-100 (IID), VGG16 | accuracy: 58.85 -> 67.52 | Improved |
| use DecFus instead of DFLAvg | DFLAvg · CIFAR-100 (alpha=0.1), ResNet-50 | accuracy: 30.55 -> 35.76 | Improved |
| use DecFus instead of DFLAvg | DFLAvg · SVHN (alpha=0.1), VGG16 | accuracy: 75.02 -> 72.84 | Hurt |
| use EDecFus (communication-efficient) instead of DecFus | DecFus · CIFAR-10 (IID), ResNet-50 | accuracy: 78.24 -> 76.93 | Hurt |
| vary hyperparameters around default (ablation) | DecFus · SVHN (alpha=1.0), VGG16 | accuracy: 93.97 -> 91.84 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DisPFL — DisPFL: Towards Communication-Efficient Personalized Federated Learning via Decentralized Sparse Training (ICML 2022) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Decentralized Sporadic Federated Learning: A Unified Algorithmic Framework with Convergence Guarantees (ICLR 2025 · similarity 0.89)
- Overcoming Data and Model heterogeneities in Decentralized Federated Learning via Synthetic Anchors (ICML 2024 · similarity 0.88)
- NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel (ICML 2025 · similarity 0.88)
- SemiDFL: A Semi-Supervised Paradigm for Decentralized Federated Learning (AAAI 2025 · similarity 0.87)
- Decentralized Directed Collaboration for Personalized Federated Learning (CVPR 2024 · similarity 0.86)
- Byzantine-robust Decentralized Federated Learning via Dual-domain Clustering and Trust Bootstrapping (CVPR 2024 · similarity 0.86)
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.
