Research map / Federated learning and privacy / Communication-efficient FL
Elastic Aggregation for Federated Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated training with ResNet-20 · CIFAR-100 | test accuracy: 32.31 -> 56.64 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated logistic regression · MNIST | test accuracy: 70.14 -> 73.64 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated training with 2Conv+2Linear · EMNIST | test accuracy: 88.71 -> 89.82 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated training with ResNet-20 · CIFAR-10 | test accuracy: 66.84 -> 68.74 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated training with ResNet-20 · CINIC-10, 200 rounds | test accuracy: 35.81 -> 36.29 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Federated training with ResNet-20 · CINIC-10, 4000 rounds | test accuracy: 68.68 -> 69.25 | Helped |
| Use parameter-sensitivity-weighted aggregation instead of naïve aggregation | Shakespeare LSTM · Shakespeare | accuracy: 62.14% -> 64.73% | Helped |
| Add elastic aggregation to an existing federated optimizer | FedAvg, FedAvgM, FedProx, SCAFFOLD, AdaOpt, PFNM, Per-FedAvg, and pFedMe · Federated optimizer experiments | performance: 3.03, 1.96, 2.12, 3.32, 2.34, 2.41, 2.80, 2.43 | Helped |
| Vary τ from 0.0 to 1.0 rather than use naïve aggregation (ablation) | Elastic aggregation · CIFAR-10 experiment in Table 1 | test accuracy: 60.74; 58.30, 59.76, 61.17, 61.91, 63.39, 62.74, 63.39, 63.23, 63.22, 60.83, 61.37 | Mixed |
| Set τ to 0.0 (ablation) | Elastic aggregation · CIFAR-10 experiment in Table 1 | test accuracy: 60.74 -> 58.30 | Hurt |
| Set τ to 0.5 (ablation) | Elastic aggregation · CIFAR-10 experiment in Table 1 | test accuracy: 60.74 -> 62.74 | Helped |
| Set τ to 0.9 (ablation) | Elastic aggregation · CIFAR-10 experiment in Table 1 | test accuracy: 60.74 -> 60.83 | Helped |
Closest papers by meaning
- FedAWA: Adaptive Optimization of Aggregation Weights in Federated Learning Using Client Vectors (CVPR 2025 · similarity 0.90)
- FedEL: Federated Elastic Learning for Heterogeneous Devices (NeurIPS 2025 · similarity 0.89)
- An Aggregation-Free Federated Learning for Tackling Data Heterogeneity (CVPR 2024 · similarity 0.88)
- FedAA: A Reinforcement Learning Perspective on Adaptive Aggregation for Fair and Robust Federated Learning (AAAI 2025 · similarity 0.88)
- Efficient Federated Learning against Heterogeneous and Non-stationary Client Unavailability (NeurIPS 2024 · similarity 0.88)
- Self-Driven Entropy Aggregation for Byzantine-Robust Heterogeneous Federated Learning (ICML 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.