Research map / Federated learning and privacy / Personalized and heterogeneous FL
FedSDR: Federated Graph Learning with Structural Noise Detection and Reconstruction
Jiaqi Liu, Zihan Tan, Guancheng Wan, Wenke Huang, He Li, Mang Ye
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 SNAA (structural noise-aware aggregation) to FedAvg (ablation) | FedSDR · PubMed | test accuracy (%): 78.41 -> 81.82 | Improved |
| add SNAA to FedAvg (ablation) | FedSDR · ogbn-products | test accuracy (%): 72.66 -> 74.25 | Improved |
| add SNAA to FedAvg (ablation) | FedSDR · Actor | test accuracy (%): 31.28 -> 31.56 | Improved |
| add RLSR (local structure reconstruction) to FedAvg (ablation) | FedSDR · ogbn-products | test accuracy (%): 72.66 -> 75.92 | Improved |
| add RLSR to FedAvg (ablation) | FedSDR · PubMed | test accuracy (%): 78.41 -> 82.26 | Improved |
| add RLSR on top of SNAA (ablation) | FedSDR · ogbn-products | test accuracy (%): 74.25 -> 78.57 | Improved |
| add SNAA on top of RLSR (ablation) | FedSDR · PubMed | test accuracy (%): 82.26 -> 82.57 | Improved |
| add SNAA on top of RLSR (ablation) | FedSDR · Actor | test accuracy (%): 32.07 -> 32.36 | Improved |
| use FedSDR instead of FedAvg under high noise | FedSDR · Roman-empire | test accuracy (%): 41.72 -> 48.33 | Improved |
| increase noise extent for FedAvg vs FedSDR | FedSDR · PubMed | test accuracy (%) | Improved |
| use heterogeneity or attack-oriented FGL methods (FedTGE, FedATH) under structural noise | FedSDR · Coauthor-CS | test accuracy (%): 82.01 (FedAvg) vs 77.54 (FedTGE) | Hurt |
| use Scaffold instead of FedAvg under structural noise | FedSDR · PubMed | test accuracy (%): 78.41 -> 41.76 | Hurt |
Closest papers by meaning
- $S^2$FGL: Spatial Spectral Federated Graph Learning (ICML 2025 · similarity 0.90)
- Subgraph Federated Learning via Spectral Methods (NeurIPS 2025 · similarity 0.90)
- Federated Graph-Level Clustering Network (AAAI 2025 · similarity 0.89)
- Decoupled Subgraph Federated Learning (ICLR 2025 · similarity 0.88)
- FedSSP: Federated Graph Learning with Spectral Knowledge and Personalized Preference (NeurIPS 2024 · similarity 0.88)
- Federated Graph Condensation with Information Bottleneck Principles (AAAI 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.
