Code for implementation of "Distributional Shapley: A Distributional Framework for Data Valuation".
Please cite the following work if you use this benchmark or the provided tools or implementations:
@inproceedings{ghorbani2020distributional,
title={A Distributional Framework for Data Valuation},
author={Ghorbani, Amirata, P. Kim, Michael and Zou, James},
booktitle={International Conference on Machine Learning},
year={2020}
}
- Python, NumPy, Tensorflow 1.12, Scikit-learn, Matplotlib
To estimate an equitbale measure of value of data points coming from an underlying distribution given a machine learning model class and a performance metric (test accuracy, etc)
This project is licensed under the MIT License - see the LICENSE.md file for details