https://github.com/openmined/pydpvalidator
Validation assets for core OpenMined libraries
https://github.com/openmined/pydpvalidator
Last synced: 3 months ago
JSON representation
Validation assets for core OpenMined libraries
- Host: GitHub
- URL: https://github.com/openmined/pydpvalidator
- Owner: OpenMined
- License: mit
- Created: 2020-06-22T09:48:59.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2020-07-14T12:51:24.000Z (over 5 years ago)
- Last Synced: 2025-06-07T00:41:04.314Z (7 months ago)
- Language: Jupyter Notebook
- Size: 135 KB
- Stars: 6
- Watchers: 7
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# PyDPValidator
This repo adapts code from https://github.com/xiyangl3/adp-estimator/ to apply the techniques in that repo's [accompanying paper, Minimax Optimal Estimation of Approximate Differential Privacy on Neighboring Databases](https://arxiv.org/abs/1905.10335) to examining [PyDP](https://github.com/OpenMined/PyDP).
## Dependencies
* To run the experiments, you need to have the following libraries installed:
1. python = 3.6
2. numpy
3. scipy
* The coefficient of best polynomial approximation are pre-computed and stored as ".mat" file. The coefficient of Chebyshev polynomials of the first kind are stored as ".npy" file.
* To get the coefficient of best polynomial approximation, you need to install Chebfun in Matlab through http://www.chebfun.org/
* To get the coefficient of Chebyshev polynomials, you need to install:
1. sympy
## Citing this work
You are encouraged to cite orginal paper for acedamic research:
```bibtex
@inproceedings{liu2019minimax,
title={Minimax Optimal Estimation of Approximate Differential Privacy on Neighboring Databases},
author={Liu, Xiyang and Oh, Sewoong},
booktitle={Advances in Neural Information Processing Systems},
pages={2414--2425},
year={2019}
}
```
## License
[MIT](https://github.com/xiyangl3/adp-estimator/blob/master/LICENSE).