Research map / Federated learning and privacy / Personalized and heterogeneous FL
One-Pass Distribution Sketch for Measuring Data Heterogeneity in Federated Learning
Zichang Liu, Zhaozhuo Xu, Benjamin Coleman, Anshumali Shrivastava
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 |
|---|---|---|---|
| retrieve the nearest-neighbor personalized model via sketch distance instead of the global model | cold-start personalized FL · MNIST cold clients | test accuracy: 0.8791 -> 0.9702 | Improved |
| retrieve the nearest-neighbor personalized model instead of the global model | cold-start personalized FL · FEMNIST cold clients | test accuracy: 0.6667 -> 0.6937 | Improved |
| retrieve the nearest-neighbor personalized model instead of the global model | cold-start personalized FL · Shakespeare cold clients | test accuracy: 0.3237 -> 0.3455 | Improved |
| local fine-tuning of global model for warm clients | FedAvg global model · MNIST warm clients | test accuracy: 0.8764 -> 0.9784 | Improved |
| local fine-tuning of global model for warm clients | FedAvg global model · FEMNIST warm clients | test accuracy: 0.6593 -> 0.7336 | Improved |
| local fine-tuning of global model for warm clients | FedAvg global model · Shakespeare warm clients | test accuracy: 0.3581 -> 0.3645 | No clear effect |
| sketch-distance client selection instead of uniform random sampling | FedAvg · MNIST, MNIST Uniform+Dirichlet, FEMNIST, Shakespeare | test accuracy (convergence) | Improved |
| use sketch-based neighbor model instead of random personalized model (ablation) | cold-start personalized FL · MNIST cold clients | test accuracy | Improved |
Closest papers by meaning
- FedGPS: Statistical Rectification Against Data Heterogeneity in Federated Learning (NeurIPS 2025 · similarity 0.89)
- Modeling Inter-Intra Heterogeneity for Graph Federated Learning (AAAI 2025 · similarity 0.87)
- Tackling Intertwined Data and Device Heterogeneities in Federated Learning with Unlimited Staleness (AAAI 2025 · similarity 0.87)
- Heterogeneity-Guided Client Sampling: Towards Fast and Efficient Non-IID Federated Learning (NeurIPS 2024 · similarity 0.87)
- One-Shot Heterogeneous Federated Learning with Local Model-Guided Diffusion Models (ICML 2025 · similarity 0.87)
- Federated Learning with Sample-level Client Drift Mitigation (AAAI 2025 · 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.
