@inproceedings{jiang2026private,
author = {Jiang, Shuli and Krichene, Walid and Mayoraz, Nicolas},
title = {Private Learning with Public Feature Conditioning},
booktitle = {ICML 2026},
year = {2026},
month = {July},
abstract = {We study differentially private (DP) regression in settings where each data sample includes public, non-sensitive features—common in applications like recommendation or advertising systems. While such label DP or DP with semi-sensitive features settings have been primarily explored in the context of classification, effective approaches for regression remain underexplored. We introduce , a conditioned variant of  that leverages the structure of public feature matrices to improve optimization under privacy constraints. Motivated by the observation that these public features often exhibit rapidly decaying spectra,  incorporates a data-driven conditioning matrix to reshape the optimization landscape and accelerate convergence. We provide convergence guarantees for convex, strongly convex and non-convex settings, and recover standard  as a special case when the conditioning matrix is the identity. We show how to construct an effective conditioning matrix for  directly from public features, enabling faster convergence than  in private linear regression, without incurring additional privacy cost. Empirically,  with this conditioning matrix consistently outperforms state-of-the-art baselines across a wide range of datasets and model architectures under label DP, demonstrating strong and robust performance in practice.},
url = {http://approjects.co.za/?big=en-us/research/publication/private-learning-with-public-feature-conditioning-2/},
}