We study the canonical statistical estimation problem of linear regression from n i.i.d.~examples under (ε,δ)-differential privacy when some response variables are adversarially corrupted. We propose a variant of the popular differentially private stochastic gradient descent (DP-SGD) algorithm with two innovations: a full-batch gradient descent to improve sample complexity and a novel adaptive clipping to guarantee robustness. When there is no adversarial corruption, this algorithm improves upon the existing state-of-the-art approach and achieves a near optimal sample complexity. Under label-corruption, this is the first efficient linear regression algorithm to guarantee both (ε,δ)-DP and robustness. Synthetic experiments confirm the superiority of our approach.
@article{arxiv.2301.13273,
title = {Near Optimal Private and Robust Linear Regression},
author = {Xiyang Liu and Prateek Jain and Weihao Kong and Sewoong Oh and Arun Sai Suggala},
journal= {arXiv preprint arXiv:2301.13273},
year = {2023}
}