Research map / Federated learning and privacy / Communication-efficient FL
Federated Reconstruction: Partially Local Federated Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Train global item parameters to support reconstruction of local user embeddings | FedRecon · MovieLens 1M, unseen users | accuracy: 40.8 -> 43.3 | Helped |
| Use 500 rather than 1 local OOV embedding (ablation) | FedRecon · Stack Overflow, 1K vocabulary | top-1 accuracy: 24.1 -> 29.6 | Helped |
| Use 500 rather than 1 local OOV embedding (ablation) | FedRecon · Stack Overflow, 5K vocabulary | top-1 accuracy: 26.2 -> 28.1 | Helped |
| Use 500 rather than 1 local OOV embedding (ablation) | FedRecon · Stack Overflow, 10K vocabulary | top-1 accuracy: 26.4 -> 27.7 | Helped |
| Use the full client dataset for both support and query instead of splitting it (ablation) | FedRecon · MovieLens 1M | accuracy: 43.3 -> 42.5 | Hurt |
| Use the full client dataset for both support and query instead of splitting it (ablation) | FedRecon · Stack Overflow, 1K vocabulary | top-1 accuracy: 29.6 -> 28.8 | Hurt |
| Update local and global parameters concurrently after reconstruction (ablation) | FedRecon · MovieLens 1M | accuracy: 43.3 -> 42.1 | Hurt |
| Update local and global parameters concurrently after reconstruction (ablation) | FedRecon · Stack Overflow, 1K vocabulary | top-1 accuracy: 29.6 -> 29.3 | Hurt |
| Use Adagrad as the server optimizer (ablation) | FedRecon · MovieLens 1M | RMSE: .907 -> .883 | Helped |
| Increase core vocabulary size from 1K to 10K while using 500 local OOV embeddings (ablation) | FedRecon · Stack Overflow | top-1 accuracy: 29.6 -> 27.7 | Hurt |
| Skip local reconstruction (ablation) | FedRecon · MovieLens 1M | accuracy: 0.0 | Hurt |
| Skip local reconstruction (ablation) | FedRecon · Stack Overflow next word prediction | relative accuracy: above 90% | Hurt |
Closest papers by meaning
- Personalized Federated Collaborative Filtering: A Variational AutoEncoder Approach (AAAI 2025 · similarity 0.90)
- FedImpro: Measuring and Improving Client Update in Federated Learning (ICLR 2024 · similarity 0.88)
- Ranking-based Client Imitation Selection for Efficient Federated Learning (ICML 2024 · similarity 0.87)
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · similarity 0.87)
- Decentralized Directed Collaboration for Personalized Federated Learning (CVPR 2024 · similarity 0.87)
- Federated Recommendation with Additive Personalization (ICLR 2024 · similarity 0.87)
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.