Understanding Clipping for Federated Learning: Convergence and Client-Level Differential Privacy
arXiv:2106.13673
Abstract
Providing privacy protection has been one of the primary motivations of Federated Learning (FL). Recently, there has been a line of work on incorporating the formal privacy notion of differential privacy with FL. To guarantee the client-level differential privacy in FL algorithms, the clients' transmitted model updates have to be clipped before adding privacy noise. Such clipping operation is substantially different from its counterpart of gradient clipping in the centralized differentially private SGD and has not been well-understood. In this paper, we first empirically demonstrate that the clipped FedAvg can perform surprisingly well even with substantial data heterogeneity when training neural networks, which is partly because the clients' updates become similar for several popular deep architectures. Based on this key observation, we provide the convergence analysis of a differential private (DP) FedAvg algorithm and highlight the relationship between clipping bias and the distribution of the clients' updates. To the best of our knowledge, this is the first work that rigorously investigates theoretical and empirical issues regarding the clipping operation in FL algorithms.
References in corpus (10)
- Deep Learning with Differential Privacy
- Federated Learning: Strategies for Improving Communication Efficiency
- Federated Learning for Mobile Keyboard Prediction
- Differentially Private Federated Learning: A Client Level Perspective
- iDLG: Improved Deep Leakage from Gradients
- Differential Privacy-enabled Federated Learning for Sensitive Health Data
- A Framework for Evaluating Gradient Leakage Attacks in Federated Learning
- Variance Reduced Local SGD with Lower Communication Complexity
- First Analysis of Local GD on Heterogeneous Data
- D2P-Fed: Differentially Private Federated Learning With Efficient Communication