ICML 2022 accepted papers on Federated learning and privacy
61 of the 1,233 papers accepted at ICML 2022 (4.9%) are on Federated learning and privacy, compared with 3.6% at ICML 2021. They fall into 4 topics.
See ICML 2022 in the interactive map
Submitting to ICML? 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
Differential privacy · 23 papers
- Understanding Clipping for Federated Learning: Convergence and Client-Level Differential Privacy (21 citations)
- Improved Rates for Differentially Private Stochastic Convex Optimization with Heavy-Tailed Data (12 citations)
- The Fundamental Price of Secure Aggregation in Differentially Private Federated Learning (11 citations)
- Robin Hood and Matthew Effects: Differential Privacy Has Disparate Impact on Synthetic Data (9 citations)
- Differentially Private Coordinate Descent for Composite Empirical Risk Minimization (7 citations)
- Differentially Private Community Detection for Stochastic Block Models (5 citations)
- Tight and Robust Private Mean Estimation with Few Users (5 citations)
- Optimal Algorithms for Mean Estimation under Local Differential Privacy (4 citations)
- FriendlyCore: Practical Differentially Private Aggregation (3 citations)
- Private Adaptive Optimization with Side information (3 citations)
- Public Data-Assisted Mirror Descent for Private Model Training (3 citations)
- Differentially Private Approximate Quantiles (2 citations)
and 11 more in the interactive map.
Communication-efficient FL · 20 papers
- DisPFL: Towards Communication-Efficient Personalized Federated Learning via Decentralized Sparse Training (45 citations)
- DAdaQuant: Doubly-adaptive quantization for communication-efficient Federated Learning (38 citations)
- Federated Learning with Partial Model Personalization (24 citations)
- Communication-Efficient Adaptive Federated Learning (23 citations)
- Accelerated Federated Learning with Decoupled Adaptive Optimization (13 citations)
- FedNL: Making Newton-Type Methods Applicable to Federated Learning (12 citations)
- ProgFed: Effective, Communication, and Computation Efficient Federated Learning by Progressive Training (12 citations)
- Personalization Improves Privacy-Accuracy Tradeoffs in Federated Learning (6 citations)
- Anarchic Federated Learning (5 citations)
- Bitwidth Heterogeneous Federated Learning with Progressive Weight Dequantization (5 citations)
- EDEN: Communication-Efficient and Robust Distributed Mean Estimation for Federated Learning (5 citations)
- Compressed-VFL: Communication-Efficient Learning with Vertically Partitioned Data (4 citations)
and 8 more in the interactive map.
Personalized and heterogeneous FL · 12 papers
- FedScale: Benchmarking Model and System Performance of Federated Learning at Scale (36 citations)
- Federated Learning with Label Distribution Skew via Logits Calibration (23 citations)
- Personalized Federated Learning through Local Memorization (21 citations)
- Generalized Federated Learning via Sharpness Aware Minimization (19 citations)
- Orchestra: Unsupervised Federated Learning via Globally Consistent Clustering (19 citations)
- Personalized Federated Learning via Variational Bayesian Inference (16 citations)
- Virtual Homogeneity Learning: Defending against Data Heterogeneity in Federated Learning (15 citations)
- Architecture Agnostic Federated Learning for Neural Networks (11 citations)
- Multi-Level Branched Regularization for Federated Learning (10 citations)
- Disentangled Federated Learning for Tackling Attributes Skew via Invariant Aggregation and Diversity Transferring (9 citations)
- Federated Learning with Positive and Unlabeled Data (5 citations)
- Deep Neural Network Fusion via Graph Matching with Applications to Model Ensemble and Federated Learning
Privacy attacks and secure learning · 6 papers
- Deduplicating Training Data Mitigates Privacy Risks in Language Models (44 citations)
- Neurotoxin: Durable Backdoors in Federated Learning (30 citations)
- Privacy for Free: How does Dataset Condensation Help Privacy? (29 citations)
- Fishing for User Data in Large-Batch Federated Learning via Gradient Magnification (14 citations)
- Selective Network Linearization for Efficient Private Inference (7 citations)
- Bounding Training Data Reconstruction in Private (Deep) Learning (4 citations)
