This repository contains the code to reproduce the experiments carried out in Individual Privacy Accounting with Gaussian Differential Privacy.
The instructions, required dependencies and the code can be found in the mimic_experiments
folder.
To ask questions or report issues, please open an issue on the issues tracker.
If you use this code, please cite our paper.
@inproceedings{koskela2023individual,
title={Individual Privacy Accounting with Gaussian Differential Privacy},
author={Antti Koskela and Marlon Tobaben and Antti Honkela},
booktitle={The Eleventh International Conference on Learning Representations},
year={2023},
url={https://openreview.net/forum?id=JmC_Tld3v-f}
}