English

DP-REC: Private & Communication-Efficient Federated Learning

Machine Learning 2021-12-09 v2 Cryptography and Security Machine Learning

Abstract

Privacy and communication efficiency are important challenges in federated training of neural networks, and combining them is still an open problem. In this work, we develop a method that unifies highly compressed communication and differential privacy (DP). We introduce a compression technique based on Relative Entropy Coding (REC) to the federated setting. With a minor modification to REC, we obtain a provably differentially private learning algorithm, DP-REC, and show how to compute its privacy guarantees. Our experiments demonstrate that DP-REC drastically reduces communication costs while providing privacy guarantees comparable to the state-of-the-art.

Keywords

Cite

@article{arxiv.2111.05454,
  title  = {DP-REC: Private & Communication-Efficient Federated Learning},
  author = {Aleksei Triastcyn and Matthias Reisser and Christos Louizos},
  journal= {arXiv preprint arXiv:2111.05454},
  year   = {2021}
}
R2 v1 2026-06-24T07:33:06.740Z