ICML 2020 accepted papers on Federated learning and privacy
22 of the 1,087 papers accepted at ICML 2020 (2.0%) are on Federated learning and privacy, compared with 2.3% at ICML 2019. They fall into 4 topics.
See ICML 2020 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 · 14 papers
- New Oracle-Efficient Algorithms for Private Synthetic Data Release (22 citations)
- Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead (18 citations)
- Optimal Differential Privacy Composition for Exponential Mechanisms (17 citations)
- Differentially Private Set Union (16 citations)
- Context Aware Local Differential Privacy (11 citations)
- Bayesian Differential Privacy for Machine Learning (9 citations)
- Data-Dependent Differentially Private Parameter Learning for Directed Graphical Models (9 citations)
- Private Query Release Assisted by Public Data (8 citations)
- Private Reinforcement Learning with PAC and Regret Guarantees (8 citations)
- Private Outsourced Bayesian Optimization (7 citations)
- Sharp Composition Bounds for Gaussian Differential Privacy via Edgeworth Expansion (6 citations)
- An end-to-end Differentially Private Latent Dirichlet Allocation Using a Spectral Algorithm (5 citations)
and 2 more in the interactive map.
Communication-efficient FL · 5 papers
- SCAFFOLD: Stochastic Controlled Averaging for Federated Learning (709 citations)
- The Non-IID Data Quagmire of Decentralized Machine Learning (88 citations)
- FedBoost: A Communication-Efficient Algorithm for Federated Learning (69 citations)
- FetchSGD: Communication-Efficient Federated Learning with Sketching (65 citations)
- From Local SGD to Local Fixed-Point Methods for Federated Learning (46 citations)
Privacy attacks and secure learning · 2 papers
- InstaHide: Instance-hiding Schemes for Private Distributed Learning (19 citations)
- Alleviating Privacy Attacks via Causal Learning (2 citations)
Personalized and heterogeneous FL · 1 paper
- Federated Learning with Only Positive Labels (42 citations)
