Minimax signal detection in regression under differential privacy
Abstract
We establish minimax separation rates across sparsity levels for regression signal detection under $(ε, δ)$-central differential privacy, considering linear regression with an unknown error distribution and generalised linear models (GLMs) with an unknown intercept. Without sparsity restrictions, the separation rates for linear regression and GLMs with binary responses match the private Gaussian mean testing rate up to polylogarithmic factors. For GLMs with more than two possible response outcomes, additional information in the marginal response distribution permits faster detection of sufficiently small signals when the dimension exceeds the sample size n. Our upper bounds are based on a reduction to private mean testing, tests of dependence between the response and sparse covariate projections, and a response-only test that is adjusted for the unknown intercept. The resulting procedure adapts to unknown sparsity and matches our lower bounds up to polylogarithmic factors. We also construct polynomial-time tests that attain these rates when ε≳ n−1/2, and compare their numerical performance with existing private procedures.
Related
- Minimax Wasserstein Testing under Differential Privacy
- Learning Sparse Support under Differential Privacy: Adaptive Algorithms and Minimax Limits
- Privacy-accuracy trade-off optimization in sensitive surveys through regression estimation
- Robust Local Fréchet Regression: Nonasymptotic Theory and Minimax Optimality
- Prophet Inequalities under Local Differential Privacy
BibTeX
@misc{hung2026minimax,
title = {{Minimax signal detection in regression under differential privacy}},
author = {Elly K. H. Hung},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.04064},
eprint = {2611.04064},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.04064},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Minimax signal detection in regression under differential privacy
AU - Elly K. H. Hung
PY - 2026
DA - 2026/11/02
PB - arXiv
DO - 10.48550/arxiv.2611.04064
UR - https://arxiv.org/abs/2611.04064
AB - We establish minimax separation rates across sparsity levels for regression signal detection under $(ε, δ)$-central differential privacy, considering linear regression with an unknown error distribution and generalised linear models (GLMs) with an unknown intercept. Without sparsity restrictions, the separation rates for linear regression and GLMs with binary responses match the private Gaussian mean testing rate up to polylogarithmic factors. For GLMs with more than two possible response outcomes, additional information in the marginal response distribution permits faster detection of sufficiently small signals when the dimension exceeds the sample size n. Our upper bounds are based on a reduction to private mean testing, tests of dependence between the response and sparse covariate projections, and a response-only test that is adjusted for the unknown intercept. The resulting procedure adapts to unknown sparsity and matches our lower bounds up to polylogarithmic factors. We also construct polynomial-time tests that attain these rates when ε≳ n^−1/2, and compare their numerical performance with existing private procedures.
ER -