ICLR 2021 accepted papers on Federated learning and privacy
17 of the 860 papers accepted at ICLR 2021 (2.0%) are on Federated learning and privacy, compared with 0.9% at ICLR 2020. They fall into 4 topics.
See ICLR 2021 in the interactive map
Submitting to ICLR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Communication-efficient FL · 6 papers
- HeteroFL: Computation and Communication Efficient Federated Learning for Heterogeneous Clients (192 citations)
- Adaptive Federated Optimization (122 citations)
- Federated Learning Based on Dynamic Regularization (114 citations)
- FedMix: Approximation of Mixup under Mean Augmented Federated Learning (66 citations)
- Federated Learning via Posterior Averaging: A New Perspective and Practical Algorithms (43 citations)
- Achieving Linear Speedup with Partial Worker Participation in Non-IID Federated Learning (14 citations)
Personalized and heterogeneous FL · 4 papers
- FedBN: Federated Learning on Non-IID Features via Local Batch Normalization (342 citations)
- FedBE: Making Bayesian Model Ensemble Applicable to Federated Learning (28 citations)
- Personalized Federated Learning with First Order Model Optimization (26 citations)
- Federated Semi-Supervised Learning with Inter-Client Consistency & Disjoint Learning (23 citations)
Differential privacy · 4 papers
- Differentially Private Learning Needs Better Features (or Much More Data) (78 citations)
- Do not Let Privacy Overbill Utility: Gradient Embedding Perturbation for Private Learning (25 citations)
- Bypassing the Ambient Dimension: Private SGD with Gradient Subspace Identification (15 citations)
- Private Post-GAN Boosting (9 citations)
Privacy attacks and secure learning · 3 papers
- CaPC Learning: Confidential and Private Collaborative Learning (28 citations)
- Information Laundering for Model Privacy (10 citations)
- R-GAP: Recursive Gradient Attack on Privacy (4 citations)
