Research map / Federated learning and privacy / Communication-efficient FL
Bias-Variance Reduced Local SGD for Less Heterogeneous Federated Learning
Tomoya Murata, Taiji Suzuki
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 BVR-L-SGD instead of the implemented competing methods | CIFAR10 classification with heterogeneous local datasets · CIFAR10, q varied | test accuracy | Improved |
| Use BVR-L-SGD instead of the implemented competing methods | CIFAR10 classification · CIFAR10, 3,000 communication rounds | train loss | Improved |
| Increase local computation budget | Local methods · CIFAR10, smallest heterogeneity | test accuracy | Improved |
| Increase local computation budget | Non-local methods · CIFAR10, smallest heterogeneity | test accuracy | No clear effect |
| Increase heterogeneity | Local SGD and SCAFFOLD · CIFAR10, q varied | train loss | Hurt |
| Increase heterogeneity | BVR-L-SGD · CIFAR10, q varied | test accuracy | Hurt |
| Use BVR-L-SGD rather than minibatch SARAH | General nonconvex smooth objectives with small second-order heterogeneity · General nonconvex smooth objectives | communication complexity | Improved |
Closest papers by meaning
- Escaping Saddle Points with Bias-Variance Reduced Local Perturbed SGD for Communication Efficient Nonconvex Distributed Learning (NeurIPS 2022 · similarity 0.94)
- SILVER: Single-loop variance reduction and application to federated learning (ICML 2024 · similarity 0.89)
- LOSCAR-SGD: Local SGD with Communication-Computation Overlap and Delay-Corrected Sparse Model Averaging (NeurIPS 2026 · similarity 0.89)
- Sharp Gaussian approximations for Decentralized Federated Learning (NeurIPS 2025 · similarity 0.88)
- Communication-Efficient Gradient Descent-Accent Methods for Distributed Variational Inequalities: Unified Analysis and Local Updates (ICLR 2024 · similarity 0.88)
- SLowcalSGD : Slow Query Points Improve Local-SGD for Stochastic Convex Optimization (NeurIPS 2024 · similarity 0.88)
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.
