Federated Generative Privacy

Aleksei Triastcyn; Boi Faltings

Federated Generative Privacy

Machine Learning 2019-10-21 v1 Cryptography and Security Distributed, Parallel, and Cluster Computing Machine Learning

Authors: Aleksei Triastcyn , Boi Faltings

Abstract

In this paper, we propose FedGP, a framework for privacy-preserving data release in the federated learning setting. We use generative adversarial networks, generator components of which are trained by FedAvg algorithm, to draw privacy-preserving artificial data samples and empirically assess the risk of information disclosure. Our experiments show that FedGP is able to generate labelled data of high quality to successfully train and validate supervised models. Finally, we demonstrate that our approach significantly reduces vulnerability of such models to model inversion attacks.

Keywords

differential privacy generative adversarial networks

Cite

@article{arxiv.1910.08385,
  title  = {Federated Generative Privacy},
  author = {Aleksei Triastcyn and Boi Faltings},
  journal= {arXiv preprint arXiv:1910.08385},
  year   = {2019}
}

Comments

IJCAI Workshop on Federated Machine Learning for User Privacy and Data Confidentiality (FL-IJCAI 2019). 7 pages, 2 figures, 3 tables

Related papers

View all related →