Optimal Unbiased Randomizers for Regression with Label Differential Privacy
Abstract
We propose a new family of label randomizers for training regression models under the constraint of label differential privacy (DP). In particular, we leverage the trade-offs between bias and variance to construct better label randomizers depending on a privately estimated prior distribution over the labels. We demonstrate that these randomizers achieve state-of-the-art privacy-utility trade-offs on several datasets, highlighting the importance of reducing bias when training neural networks with label DP. We also provide theoretical results shedding light on the structural properties of the optimal unbiased randomizers.
Cite
Text
Varadaraja et al. "Optimal Unbiased Randomizers for Regression with Label Differential Privacy." Neural Information Processing Systems, 2023.Markdown
[Varadaraja et al. "Optimal Unbiased Randomizers for Regression with Label Differential Privacy." Neural Information Processing Systems, 2023.](https://mlanthology.org/neurips/2023/varadaraja2023neurips-optimal/)BibTeX
@inproceedings{varadaraja2023neurips-optimal,
title = {{Optimal Unbiased Randomizers for Regression with Label Differential Privacy}},
author = {Varadaraja, Ashwinkumar Badanidiyuru and Ghazi, Badih and Kamath, Pritish and Kumar, Ravi and Leeman, Ethan and Manurangsi, Pasin and Varadarajan, Avinash V and Zhang, Chiyuan},
booktitle = {Neural Information Processing Systems},
year = {2023},
url = {https://mlanthology.org/neurips/2023/varadaraja2023neurips-optimal/}
}