weijiaheng / credible-sample-elicitation

[AISTATS2021] Official implementation of "Sample Elicitation"

Sample elicitation

This code is the official implementation of our paper "Sample Elicitation" accepted by AISTATS2021.

Utilities

📋 (1) Syntheric Experiment: this folder is the Matlab implementation of our porposed mechanism w.r.t. synthetic data (as an example, we adopt the synthetic data drawn from 2-D Gaussian distributions).

📋 (2) MNIST Experiment: this folder is the Python implementation of our porposed mechanism given image data (as an example, we adopt MNIST test dataset).

Note: details of reproducing our experiment results are mentioned in the ``README" file in each folder.

Citation

If you use our code, please cite the following paper:

@inproceedings{wei2021sample,
  title={Sample Elicitation},
  author={Wei, Jiaheng and Fu, Zuyue and Liu, Yang and Li, Xingyu and Yang, Zhuoran and Wang, Zhaoran},
  booktitle={International Conference on Artificial Intelligence and Statistics},
  pages={2692--2700},
  year={2021},
  organization={PMLR}
}

Contact

If you have any concerns on the impletation, feel free to send me an e-mail ([email protected]) or report an issue.

References

The codes of fid.py, fid_peer.py are based on https://github.com/mseitzer/pytorch-fid/blob/master/pytorch_fid

Thanks for watching!