Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.21959 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910272715751424 |
|---|---|
| author | Chen, Yu-Wei Pasupathy, Raghu Awan, Jordan |
| author_facet | Chen, Yu-Wei Pasupathy, Raghu Awan, Jordan |
| contents | We develop a near-optimal testing procedure under the framework of Gaussian differential privacy for simple as well as one- and two-sided tests under monotone likelihood ratio conditions. Our mechanism is based on a private mean estimator with data-driven clamping bounds, whose population risk matches the private minimax rate up to logarithmic factors. Using this estimator, we construct private test statistics that achieve the same asymptotic relative efficiency as the non-private, most powerful tests while maintaining conservative type I error control. In addition to our theoretical results, our numerical experiments show that our private tests outperform competing DP methods and offer comparable power to the non-private most powerful tests, even at moderately small sample sizes and privacy loss budgets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_21959 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Near-Optimal Private Tests for Simple and MLR Hypotheses Chen, Yu-Wei Pasupathy, Raghu Awan, Jordan Machine Learning We develop a near-optimal testing procedure under the framework of Gaussian differential privacy for simple as well as one- and two-sided tests under monotone likelihood ratio conditions. Our mechanism is based on a private mean estimator with data-driven clamping bounds, whose population risk matches the private minimax rate up to logarithmic factors. Using this estimator, we construct private test statistics that achieve the same asymptotic relative efficiency as the non-private, most powerful tests while maintaining conservative type I error control. In addition to our theoretical results, our numerical experiments show that our private tests outperform competing DP methods and offer comparable power to the non-private most powerful tests, even at moderately small sample sizes and privacy loss budgets. |
| title | Near-Optimal Private Tests for Simple and MLR Hypotheses |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.21959 |