Research map / Federated learning and privacy / Personalized and heterogeneous FL
A Coupled Design of Exploiting Record Similarity for Practical Vertical Federated Learning
Zhaomin Wu, Qinbin Li, Bingsheng He
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 |
|---|---|---|---|
| remove sort gate (ablation) | FedSim · house | RMSE: 42.12 -> 52.14 | Hurt |
| remove sort gate (ablation) | FedSim · hdb | RMSE: 27.13 -> 36.35 | Hurt |
| remove sort gate (ablation) | FedSim · company | RMSE: 37.08 -> 38.28 | Hurt |
| remove weight gate (ablation) | FedSim · company | RMSE: 37.08 -> 41.00 | Hurt |
| remove weight gate (ablation) | FedSim · house | RMSE: 42.12 -> 42.82 | Hurt |
| remove CNN in merge model (ablation) | FedSim · company | RMSE: 37.08 -> 39.63 | Hurt |
| remove CNN in merge model (ablation) | FedSim · game | accuracy (%): 92.88 -> 92.50 | Hurt |
| remove weight gate (ablation) | FedSim · game | accuracy (%): 92.88 -> 92.79 | No clear effect |
| use FedSim soft multi-record linkage instead of one-to-one Top1Sim linkage | SplitNN · bike | RMSE: 256.19 -> 235.67 | Improved |
| use FedSim instead of AvgSim averaging over K neighbors | SplitNN · game | accuracy (%): 90.84 -> 92.88 | Improved |
| append similarity as feature (FeatureSim) instead of FedSim | SplitNN · house | RMSE: 58.31 -> 66.39 | Hurt |
| link with FedSim under identifier noise sigma_cf=0.2 | FedSim vs Top1Sim · frog | accuracy (%): 84% -> 91% | Improved |
Closest papers by meaning
- Federated Transformer: Multi-Party Vertical Federated Learning on Practical Fuzzily Linked Data (NeurIPS 2024 · similarity 0.88)
- VFLAIR: A Research Library and Benchmark for Vertical Federated Learning (ICLR 2024 · similarity 0.87)
- FedWon: Triumphing Multi-domain Federated Learning Without Normalization (ICLR 2024 · similarity 0.87)
- PubSub-VFL: Towards Efficient Two-Party Split Learning in Heterogeneous Environments via Publisher/Subscriber Architecture (NeurIPS 2025 · similarity 0.87)
- Vertical Federated Learning with Missing Features During Training and Inference (ICLR 2025 · similarity 0.86)
- Complementary Knowledge Distillation for Robust and Privacy-Preserving Model Serving in Vertical Federated Learning (AAAI 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.
