Research map / Federated learning and privacy / Communication-efficient FL
Distributed Personalized Empirical Risk Minimization
Yuyang Deng, Mohammad Mahdi Kamani, Pouria Mahdavinia, Mehrdad Mahdavi
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 PERM single-loop personalization instead of localized FedAvg, perFedAvg and pFedME | PERM · synthetic data, 50 clients | personalized accuracy and loss | Improved |
| use PERM instead of PerFedAvg, pFedMe and localized FedAvg after 10 warm-up rounds | PERM · CIFAR10, 2-layer CNN, 2 classes per client | personalized validation convergence | Improved |
| 10 communication rounds of warm-up before personalization in PERM | PERM · CIFAR10 | initial personalized validation | Hurt |
| use PERM instead of localized FedAvg and perFedAvg | PERM · EMNIST, 50 clients, 2-layer MLP | personalized accuracy and loss | Improved |
| estimate mixing parameters in PERM compared with FedAvg per communication round | PERM single loop · wall-clock per round, 20 local steps | run time | Hurt |
| homogeneous vs highly heterogeneous (1 class per client) data for learned mixing weights alpha | two-stage PERM · MNIST, 50 clients, MLP | sparsity of learned alpha | Mixed |
Closest papers by meaning
- Minimax Estimation for Personalized Federated Learning: An Alternative between FedAvg and Local Training? (JMLR 2023 · similarity 0.88)
- On the Asymptotic Distribution of the Minimum Empirical Risk (ICML 2024 · similarity 0.86)
- Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees (ICML 2025 · similarity 0.85)
- PerAda: Parameter-Efficient Federated Learning Personalization with Generalization Guarantees (CVPR 2024 · similarity 0.85)
- A Bayesian Approach for Personalized Federated Learning in Heterogeneous Settings (NeurIPS 2024 · similarity 0.85)
- Harnessing Heterogeneous Statistical Strength for Personalized Federated Learning via Hierarchical Bayesian Inference (ICML 2025 · similarity 0.84)
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.
