Skip to yearly menu bar Skip to main content


Optimal Unbiased Randomizers for Regression with Label Differential Privacy

Ashwinkumar Badanidiyuru Varadaraja · Badih Ghazi · Pritish Kamath · Ravi Kumar · Ethan Leeman · Pasin Manurangsi · Avinash V Varadarajan · Chiyuan Zhang

Great Hall & Hall B1+B2 (level 1) #1612
[ ]
[ Paper [ Poster [ OpenReview
Wed 13 Dec 3 p.m. PST — 5 p.m. PST


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.

Chat is not available.