Paper 2022/933
Secure Quantized Training for Deep Learning
, CSIRO's Data61, CSIRO's Data61Abstract
We implement training of neural networks in secure multi-party computation (MPC) using quantization commonly used in said setting. We are the first to present an MNIST classifier purely trained in MPC that comes within 0.2 percent of the accuracy of the same convolutional neural network trained via plaintext computation. More concretely, we have trained a network with two convolutional and two dense layers to 99.2% accuracy in 3.5 hours (under one hour for 99% accuracy). We have also implemented AlexNet for CIFAR-10, which converges in a few hours. We develop novel protocols for exponentiation and inverse square root. Finally, we present experiments in a range of MPC security models for up to ten parties, both with honest and dishonest majority as well as semi-honest and malicious security.
Metadata
- Available format(s)
-
PDF
- Category
- Implementation
- Publication info
- Published elsewhere. International Conference on Machine Learning
- Keywords
- Privacy-preserving machine learning secure multi-party computation
- Contact author(s)
-
mks keller @ gmail com
ke sun @ data61 csiro au - History
- 2022-07-18: approved
- 2022-07-18: received
- See all versions
- Short URL
- https://ia.cr/2022/933
- License
-
CC BY-NC-SA
BibTeX
@misc{cryptoeprint:2022/933,
author = {Marcel Keller and Ke Sun},
title = {Secure Quantized Training for Deep Learning},
howpublished = {Cryptology {ePrint} Archive, Paper 2022/933},
year = {2022},
url = {https://eprint.iacr.org/2022/933}
}
