Stochastic-Sign SGD for Federated Learning with Theoretical Guarantees
This addresses the need for robust and efficient parameter estimation in federated learning, particularly for heterogeneous data settings, though it appears incremental as it builds on SIGNSGD.
The paper tackles the challenge of developing a federated learning method that is communication efficient, differentially private, and Byzantine resilient with theoretical convergence guarantees, proposing Stochastic-Sign SGD, which shows effectiveness in experiments on MNIST and CIFAR-10 datasets.
Federated learning (FL) has emerged as a prominent distributed learning paradigm. FL entails some pressing needs for developing novel parameter estimation approaches with theoretical guarantees of convergence, which are also communication efficient, differentially private and Byzantine resilient in the heterogeneous data distribution settings. Quantization-based SGD solvers have been widely adopted in FL and the recently proposed SIGNSGD with majority vote shows a promising direction. However, no existing methods enjoy all the aforementioned properties. In this paper, we propose an intuitively-simple yet theoretically-sound method based on SIGNSGD to bridge the gap. We present Stochastic-Sign SGD which utilizes novel stochastic-sign based gradient compressors enabling the aforementioned properties in a unified framework. We also present an error-feedback variant of the proposed Stochastic-Sign SGD which further improves the learning performance in FL. We test the proposed method with extensive experiments using deep neural networks on the MNIST dataset and the CIFAR-10 dataset. The experimental results corroborate the effectiveness of the proposed method.