We present one-shot federated learning, where a central server learns a global model over a network of federated devices in a single round of communication. Our approach - drawing on ensemble learning and knowledge aggregation - achieves an average relative gain of 51.5% in AUC over local baselines and comes within 90.1% of the (unattainable) global ideal. We discuss these methods and identify several promising directions of future work.
@article{arxiv.1902.11175,
title = {One-Shot Federated Learning},
author = {Neel Guha and Ameet Talwalkar and Virginia Smith},
journal= {arXiv preprint arXiv:1902.11175},
year = {2019}
}
Comments
5 pages, 3 figures, 1 table. 2nd Workshop on Machine Learning on the Phone and other Consumer Devices, NeurIPs 2018