Federated Two-stage Learning with Sign-based Voting
This work addresses communication efficiency and privacy in federated learning for distributed systems, though it appears incremental as it builds on existing methods like sign-based SGD and cut layers.
The paper tackles communication bottlenecks and deployment difficulties in federated learning by proposing a two-stage framework with a cut layer for local representation learning and sign-based SGD with majority voting for model updates, resulting in reduced global parameters and alleviated communication limitations.
Federated learning is a distributed machine learning mechanism where local devices collaboratively train a shared global model under the orchestration of a central server, while keeping all private data decentralized. In the system, model parameters and its updates are transmitted instead of raw data, and thus the communication bottleneck has become a key challenge. Besides, recent larger and deeper machine learning models also pose more difficulties in deploying them in a federated environment. In this paper, we design a federated two-stage learning framework that augments prototypical federated learning with a cut layer on devices and uses sign-based stochastic gradient descent with the majority vote method on model updates. Cut layer on devices learns informative and low-dimension representations of raw data locally, which helps reduce global model parameters and prevents data leakage. Sign-based SGD with the majority vote method for model updates also helps alleviate communication limitations. Empirically, we show that our system is an efficient and privacy preserving federated learning scheme and suits for general application scenarios.