Research map / Optimization / Distributed and compressed
Distributionally Robust Federated Averaging
Yuyang Deng, Mohammad Mahdi Kamani, Mehrdad Mahdavi
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 |
|---|---|---|---|
| increase synchronization gap tau from 5 to 15 in DRFA (ablation) | DRFA · Fashion MNIST logistic regression, 10 devices | communication rounds to reach 50% worst-distribution accuracy | Improved |
| increase synchronization gap tau from 5 to 15 in DRFA (ablation) | DRFA · Fashion MNIST logistic regression, 10 devices | local iterations to reach 50% worst-distribution accuracy | Hurt |
| use DRFA (periodic lambda updates) instead of FedAvg | federated logistic regression · Fashion MNIST, worst-distribution accuracy | rounds to 50% worst accuracy: FedAvg cannot reach 50% after 300 rounds | Improved |
| use DRFA instead of AFL (tau=1, lambda updated every iteration) | AFL · Fashion MNIST, tau=10 | communication rounds to equal worst accuracy | Improved |
| use DRFA instead of AFL | AFL · Fashion MNIST | local iterations to reach 50% worst accuracy | Hurt |
| use DRFA instead of FedAvg | FedAvg · Shakespeare, worst-distribution accuracy | worst-distribution accuracy | No clear effect |
| use DRFA instead of FedAvg | FedAvg · Fashion MNIST, global training accuracy over 100 rounds | average training accuracy | No clear effect |
Closest papers by meaning
- SCAFFOLD: Stochastic Controlled Averaging for Federated Learning (ICML 2020 · similarity 0.88)
- Decentralized Sporadic Federated Learning: A Unified Algorithmic Framework with Convergence Guarantees (ICLR 2025 · similarity 0.88)
- Stochastic Controlled Averaging for Federated Learning with Communication Compression (ICLR 2024 · similarity 0.87)
- Federated Generalised Variational Inference: A Robust Probabilistic Federated Learning Framework (ICML 2025 · similarity 0.86)
- Communication-Efficient Federated Group Distributionally Robust Optimization (NeurIPS 2024 · similarity 0.86)
- Communication-Efficient Gradient Descent-Accent Methods for Distributed Variational Inequalities: Unified Analysis and Local Updates (ICLR 2024 · similarity 0.86)
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.
