Adaptive Private-K-Selection with Adaptive K and Application to Multi-label PATE
Abstract
We provide an end-to-end Renyi DP based-framework for differentially private top-$k$ selection. Unlike previous approaches, which require a data-independent choice on $k$, we propose to privately release a data-dependent choice of $k$ such that the gap between $k$-th and the $(k+1)$st “quality” is large. This is achieved by an extension of the Report-Noisy-Max algorithm with a more concentrated Gaussian noise. Not only does this eliminates one hyperparameter, the adaptive choice of $k$ also certifies the stability of the top-$k$ indices in the unordered set so we can release them using a combination of the propose-test-release (PTR) framework and the Distance-to-Stability mechanism. We show that our construction improves the privacy-utility trade-offs compared to the previous top-$k$ selection algorithms theoretically and empirically. Additionally, we apply our algorithm to “Private Aggregation of Teacher Ensembles (PATE)” in multi-label classification tasks with a large number of labels and show that it leads to significant performance gains.
BibTeX
@InProceedings{pmlr-v151-zhu22e,
title = { Adaptive Private-K-Selection with Adaptive K and Application to Multi-label PATE },
author = {Zhu, Yuqing and Wang, Yu-Xiang},
booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics},
pages = {5622--5635},
year = {2022},
editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel},
volume = {151},
series = {Proceedings of Machine Learning Research},
month = {28--30 Mar},
publisher = {PMLR},
pdf = {https://proceedings.mlr.press/v151/zhu22e/zhu22e.pdf},
url = {https://proceedings.mlr.press/v151/zhu22e.html},
abstract = { We provide an end-to-end Renyi DP based-framework for differentially private top-$k$ selection. Unlike previous approaches, which require a data-independent choice on $k$, we propose to privately release a data-dependent choice of $k$ such that the gap between $k$-th and the $(k+1)$st “quality” is large. This is achieved by an extension of the Report-Noisy-Max algorithm with a more concentrated Gaussian noise. Not only does this eliminates one hyperparameter, the adaptive choice of $k$ also certifies the stability of the top-$k$ indices in the unordered set so we can release them using a combination of the propose-test-release (PTR) framework and the Distance-to-Stability mechanism. We show that our construction improves the privacy-utility trade-offs compared to the previous top-$k$ selection algorithms theoretically and empirically. Additionally, we apply our algorithm to “Private Aggregation of Teacher Ensembles (PATE)” in multi-label classification tasks with a large number of labels and show that it leads to significant performance gains. }
}