Research map / Federated learning and privacy / Communication-efficient FL
Convergence Analysis of Sequential Federated Learning on Heterogeneous Data
Yipeng Li, Xinchen Lyu
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 sequential FL instead of parallel FL on extremely heterogeneous data (C=1) | VGG-9 on CIFAR-10 · CIFAR-10 C=1, K=5 | test accuracy: 67.61 -> 78.43 | Improved |
| use sequential FL instead of parallel FL on extremely heterogeneous data (C=1) | ResNet-18 on CIFAR-10 · CIFAR-10 C=1, K=50 | test accuracy: 34.29 -> 68.91 | Improved |
| use sequential FL instead of parallel FL on CINIC-10 (extremely heterogeneous) | ResNet-18 on CINIC-10 · CINIC-10 C=1, K=5 | test accuracy: 41.12 -> 60.36 | Improved |
| use sequential FL instead of parallel FL on moderately heterogeneous data (C=2) | ResNet-18 on CIFAR-10 · CIFAR-10 C=2, K=50 | test accuracy: 79.88 -> 79.38 | No clear effect |
| use sequential FL instead of parallel FL on moderately heterogeneous data (C=2) | VGG-9 on CIFAR-10 · CIFAR-10 C=2, K=50 | test accuracy: 78.01 -> 79.67 | No clear effect |
| increase local steps K from 1 to 5 in SFL (ablation) | SFL VGG-9 on CIFAR-10 · CIFAR-10 C=1 and C=2 | test accuracy | Improved |
| increase local steps K beyond 5 in SFL (ablation) | SFL VGG-9 on CIFAR-10 · CIFAR-10 | test accuracy: 78.43 -> 68.86 (C=1, K=5 -> K=50) | Hurt |
| increase local steps K from 5 to 50 in PFL (ablation) | PFL VGG-9 on CIFAR-10 · CIFAR-10 C=1 | test accuracy: 67.61 -> 45.77 | Hurt |
| increase heterogeneity from C=2 to C=1 for SFL | SFL VGG-9 on CIFAR-10 · CIFAR-10 K=5 | test accuracy: 82.56 -> 78.43 | Hurt |
| 1-D quadratic with moderate heterogeneity (zeta*=1, delta=0), SFL vs PFL | SFL · Simulated quadratics Group 2 | distance to optimum | Hurt |
Closest papers by meaning
- Convergence Analysis of Split Federated Learning on Heterogeneous Data (NeurIPS 2024 · similarity 0.95)
- A New Theoretical Perspective on Data Heterogeneity in Federated Optimization (ICML 2024 · similarity 0.90)
- Exact and Linear Convergence for Federated Learning under Arbitrary Client Participation is Attainable (NeurIPS 2025 · similarity 0.89)
- FedQS: Optimizing Gradient and Model Aggregation for Semi-Asynchronous Federated Learning (NeurIPS 2025 · similarity 0.89)
- SemiDFL: A Semi-Supervised Paradigm for Decentralized Federated Learning (AAAI 2025 · similarity 0.89)
- FedLPS: Heterogeneous Federated Learning for Multiple Tasks with Local Parameter Sharing (AAAI 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.
