Research map / Federated learning and privacy / Personalized and heterogeneous FL
Reducing Training Time in Cross-Silo Federated Learning Using Multigraph Topology
Tuong Do, Binh X. Nguyen, Vuong Pham, Toan Tran, Erman Tjiputra, Quang D. Tran, Anh Nguyen
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 |
|---|---|---|---|
| replace RING overlay with multigraph topology with isolated nodes | RING · FEMNIST, Exodus network | cycle time (ms): 24.7 -> 12.1 | Improved |
| replace RING overlay with multigraph topology | RING · iNaturalist | cycle time: 1.5 times average reduction | Improved |
| replace RING overlay with multigraph topology | RING · Sentiment140 | cycle time: 1.74 times average reduction | Improved |
| replace RING with multigraph on Amazon network | RING · iNaturalist, Amazon network | cycle time (ms): 81.3 -> 81.3 | No clear effect |
| replace RING with multigraph topology | RING · FEMNIST, Exodus network | accuracy (%): 71.05 -> 71.13 | No clear effect |
| randomly remove 20 silos from RING overlay (ablation) | RING · FEMNIST, Exodus network | accuracy (%): 71.05 -> 61.2 | Hurt |
| remove the 20 most inefficient silos from RING overlay (ablation) | RING · FEMNIST, Exodus network | accuracy (%): 71.05 -> 61.48 | Hurt |
| remove 1 silo randomly from RING overlay (ablation) | RING · FEMNIST, Exodus network | accuracy (%): 71.05 -> 70.63 | Hurt |
| raise max edges t from 1 to 3 (ablation) | Multigraph · FEMNIST, Exodus network | cycle time (ms): 24.7 -> 13.5 | Improved |
| raise max edges t from 3 to 5 (ablation) | Multigraph · FEMNIST, Exodus network | cycle time (ms): 13.5 -> 12.1 | Improved |
| raise max edges t from 5 to 8 (ablation) | Multigraph · FEMNIST, Exodus network | accuracy (%): 71.13 -> 69.27 | Hurt |
| raise max edges t beyond 8 (10, 20, 30) (ablation) | Multigraph · FEMNIST, Exodus network | cycle time (ms): 11.9 -> 11.9 | No clear effect |
Closest papers by meaning
- Less is More: Federated Graph Learning with Alleviating Topology Heterogeneity from A Causal Perspective (ICML 2025 · similarity 0.86)
- Subgraph Federated Learning for Local Generalization (ICLR 2025 · similarity 0.86)
- FedCompass: Efficient Cross-Silo Federated Learning on Heterogeneous Client Devices Using a Computing Power-Aware Scheduler (ICLR 2024 · similarity 0.86)
- Modeling Inter-Intra Heterogeneity for Graph Federated Learning (AAAI 2025 · similarity 0.86)
- Personalized Subgraph Federated Learning with Differentiable Auxiliary Projections (NeurIPS 2025 · similarity 0.85)
- Virtual Nodes Can Help: Tackling Distribution Shifts in Federated Graph Learning (AAAI 2025 · 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.
