Research map / Federated learning and privacy / Communication-efficient FL
Acceleration of Federated Learning with Alleviated Forgetting in Local Training
Chencheng Xu, Zhiwei Hong, Minlie Huang, Tao Jiang
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 |
|---|---|---|---|
| Regularize local training with pseudo and perturbed data to alleviate forgetting | FedReg · MNIST (two classes) | R_0.9 rounds: 37 -> 25 | Improved |
| Regularize local training with pseudo and perturbed data to alleviate forgetting | FedReg · MNIST (one class) | R_0.9 rounds: 74 -> 32 | Improved |
| Regularize local training with pseudo and perturbed data to alleviate forgetting | FedReg · EMNIST | R_0.9 rounds: 170 -> 24 | Improved |
| Regularize local training with pseudo and perturbed data to alleviate forgetting | FedReg · CIFAR-10 (uniform) | accuracy: 0.576 -> 0.715 | Improved |
| Use pseudo-data regularization under one-class client heterogeneity | FedReg · CIFAR-10 (one class) | accuracy: 0.371 -> 0.616 | Improved |
| Use pseudo-data regularization with a ResNet-9 | FedReg · CIFAR-100 (ResNet-9) | accuracy: 0.323 -> 0.502 | Improved |
| Use pseudo-data regularization while fine-tuning a transformer | FedReg · CIFAR-100 (Transformer) | R_0.9 rounds: 21 -> 14 | Improved |
| Use pseudo-data regularization for medical image classification | FedReg · COVID-19 CT images | accuracy: 0.614 -> 0.673 | Improved |
| Use modified gradients rather than DPSGD for privacy protection | FedReg with MG · EMNIST | accuracy: 0.882 -> 0.986 | Improved |
| Use modified gradients rather than DPSGD for privacy protection | FedReg with MG · CIFAR-10 (uniform) | accuracy: 0.391 -> 0.703 | Improved |
| Increase pseudo-data generation step size from too-small values to a tuned intermediate value | FedReg · EMNIST and CIFAR-10 (one class) | communication rounds to target accuracy | Mixed |
| Add perturbed data regularization alongside pseudo data | FedReg · Local training across evaluated datasets | accuracy | Improved |
Closest papers by meaning
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · similarity 0.88)
- Federated Orthogonal Training: Mitigating Global Catastrophic Forgetting in Continual Federated Learning (ICLR 2024 · similarity 0.88)
- An Efficient and Accurate Dynamic Sparse Training Framework Based on Parameter-Freezing (AAAI 2025 · similarity 0.87)
- Federated Unlearning with Gradient Descent and Conflict Mitigation (AAAI 2025 · similarity 0.87)
- One Arrow, Two Hawks: Sharpness-aware Minimization for Federated Learning via Global Model Trajectory (ICML 2025 · similarity 0.87)
- FedLMT: Tackling System Heterogeneity of Federated Learning via Low-Rank Model Training with Theoretical Guarantees (ICML 2024 · 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.
