ICML 2023 accepted papers on Federated learning and privacy

Differential privacy · 37 papers

• Bayesian Estimation of Differential Privacy (9 citations)
• From Robustness to Privacy and Back (6 citations)
• Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning (6 citations)
• Differentially Private Optimization on Large Model at Small Cost (5 citations)
• Fully-Adaptive Composition in Differential Privacy (5 citations)
• Differential Privacy has Bounded Impact on Fairness in Classification (3 citations)
• The Price of Differential Privacy under Continual Observation (3 citations)
• The Test of Tests: A Framework for Differentially Private Hypothesis Testing (3 citations)
• Federated Linear Contextual Bandits with User-level Differential Privacy (2 citations)
• Generating Private Synthetic Data with Genetic Algorithms (2 citations)
• On the Privacy-Robustness-Utility Trilemma in Distributed Learning (2 citations)
• TAN Without a Burn: Scaling Laws of DP-SGD (2 citations)

and 25 more in the interactive map.

Personalized and heterogeneous FL · 21 papers

• FedDisco: Federated Learning with Discrepancy-Aware Collaboration (25 citations)
• Revisiting Weighted Aggregation in Federated Learning with Neural Networks (18 citations)
• Personalized Subgraph Federated Learning (11 citations)
• Efficient Personalized Federated Learning via Sparse Model-Adaptation (8 citations)
• Dynamic Regularized Sharpness Aware Minimization in Federated Learning: Approaching Global Consistency and Smooth Landscape (7 citations)
• Personalized Federated Learning under Mixture of Distributions (6 citations)
• FedVS: Straggler-Resilient and Privacy-Preserving Vertical Federated Learning for Split Models (5 citations)
• Vertical Federated Graph Neural Network for Recommender System (4 citations)
• LESS-VFL: Communication-Efficient Feature Selection for Vertical Federated Learning (3 citations)
• Optimizing the Collaboration Structure in Cross-Silo Federated Learning (3 citations)
• Conformal Prediction for Federated Uncertainty Quantification Under Label Shift (1 citation)
• Federated Conformal Predictors for Distributed Uncertainty Quantification (1 citation)

and 9 more in the interactive map.

Communication-efficient FL · 20 papers

• Fed-CBS: A Heterogeneity-Aware Client Sampling Mechanism for Federated Learning via Class-Imbalance Reduction (13 citations)
• A General Theory for Federated Optimization with Asynchronous and Heterogeneous Clients Updates (11 citations)
• Improving the Model Consistency of Decentralized Federated Learning (10 citations)
• FedBR: Improving Federated Learning on Heterogeneous Data via Local Learning Bias Reduction (8 citations)
• Federated Adversarial Learning: A Framework with Convergence Analysis (7 citations)
• On the Convergence of Federated Averaging with Cyclic Client Participation (7 citations)
• FedHPO-Bench: A Benchmark Suite for Federated Hyperparameter Optimization (6 citations)
• The Blessing of Heterogeneity in Federated Q-Learning: Linear Speedup and Beyond (4 citations)
• Sketching for First Order Method: Efficient Algorithm for Low-Bandwidth Channel and Vulnerability (2 citations)
• Achieving Linear Speedup in Non-IID Federated Bilevel Learning (1 citation)
• Anchor Sampling for Federated Learning with Partial Client Participation (1 citation)
• Byzantine-Robust Learning on Heterogeneous Data via Gradient Splitting (1 citation)

and 8 more in the interactive map.

Privacy attacks and secure learning · 12 papers

• HETAL: Efficient Privacy-preserving Transfer Learning with Homomorphic Encryption (9 citations)
• Cocktail Party Attack: Breaking Aggregation-Based Privacy in Federated Learning Using Independent Component Analysis (5 citations)
• Analyzing Privacy Leakage in Machine Learning via Multiple Hypothesis Testing: A Lesson From Fano (3 citations)
• Chameleon: Adapting to Peer Images for Planting Durable Backdoors in Federated Learning (3 citations)
• Fair yet Asymptotically Equal Collaborative Learning (2 citations)
• Surrogate Model Extension (SME): A Fast and Accurate Weight Update Attack on Federated Learning (2 citations)
• TabLeak: Tabular Data Leakage in Federated Learning (2 citations)
• Forget Unlearning: Towards True Data-Deletion in Machine Learning (1 citation)
• SRATTA: Sample Re-ATTribution Attack of Secure Aggregation in Federated Learning. (1 citation)
• GuardHFL: Privacy Guardian for Heterogeneous Federated Learning
• LeadFL: Client Self-Defense against Model Poisoning in Federated Learning
• Secure Federated Correlation Test and Entropy Estimation