Research map / Federated learning and privacy / Communication-efficient FL
Personalized Federated Learning towards Communication Efficiency, Robustness and Fairness
Shiyun Lin, Yuze Han, Xiang Li, Zhihua Zhang
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 |
|---|---|---|---|
| use lp-proj instead of FedAvg under a fixed byte budget | FedAvg · Synthetic(0,0) | test accuracy: 0.625 -> 0.885 | Improved |
| use lp-proj instead of FedAvg under a fixed byte budget | FedAvg · EMNIST | test accuracy: 0.906 | Improved |
| use lp-proj-1 instead of FedAvg to reach target accuracy | FedAvg · Synthetic(0,0) | bytes used: 597800 -> 4620 | Improved |
| use lp-proj-1 instead of FedAvg to reach target accuracy | FedAvg · EMNIST | bytes used: 445851400 -> 174720 | Improved |
| use lp-proj instead of gradient compression (QSGD) | QSGD · Synthetic(0,0) | test accuracy: 0.115 -> 0.885 | Improved |
| use lp-proj instead of Sketch | Sketch · Synthetic(0,0) | test accuracy: 0.456 -> 0.885 | Improved |
| lp-proj-2 (L2) instead of lp-proj-1 (L1) regulariser | lp-proj · EMNIST | bytes used: 174720 -> 196560 | Hurt |
| lp-proj-2 (L2) instead of lp-proj-1 (L1) regulariser | lp-proj · Synthetic(0,0) | test accuracy: 0.885 -> 0.888 | No clear effect |
| lp-proj-1 versus RSA on fairness | RSA · CIFAR | test accuracy: 77.68% -> 79.22% | Improved |
| lp-proj-2 versus Ditto on fairness | Ditto · EMNIST | test accuracy: 90.89% -> 90.70% | No clear effect |
| increase fraction of malicious clients under weak attacks (same-value, Gaussian) | lp-proj · EMNIST | test accuracy | No clear effect |
| increase fraction of malicious clients under strong attacks (sign-flipping, data poisoning) | lp-proj · CIFAR | test accuracy | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- Ditto — Ditto: Fair and Robust Federated Learning Through Personalization (ICML 2021) · better on 0 of 2 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
- A Random Projection Approach to Personalized Federated Learning: Enhancing Communication Efficiency, Robustness, and Fairness (JMLR 2024 · similarity 0.97)
- Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees (ICML 2025 · similarity 0.90)
- Fine-Tuning Personalization in Federated Learning to Mitigate Adversarial Clients (NeurIPS 2024 · similarity 0.90)
- Flexible Sharpness-Aware Personalized Federated Learning (AAAI 2025 · similarity 0.89)
- Adaptive Latent-Space Constraints in Personalized Federated Learning (NeurIPS 2025 · similarity 0.89)
- PeFLL: Personalized Federated Learning by Learning to Learn (ICLR 2024 · similarity 0.89)
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.
