Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.22320 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917478690455552 |
|---|---|
| author | Kalinin, Nikita P. Najar, Ali Roth, Valentin Lampert, Christoph H. |
| author_facet | Kalinin, Nikita P. Najar, Ali Roth, Valentin Lampert, Christoph H. |
| contents | We study continual mean estimation, where data vectors arrive sequentially and the goal is to maintain accurate estimates of the running mean. We address this problem under user-level differential privacy, which protects each user's entire dataset even when they contribute multiple data points. Previous work on this problem has focused on pure differential privacy. While important, this approach limits applicability, as it leads to overly noisy estimates. In contrast, we analyze the problem under approximate differential privacy, adopting recent advances in the Matrix Factorization mechanism. We introduce a novel mean estimation specific factorization, which is both efficient and accurate, achieving asymptotically lower mean-squared error bounds in continual mean estimation under user-level differential privacy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22320 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Matrix Factorization for Practical Continual Mean Estimation Under User-Level Differential Privacy Kalinin, Nikita P. Najar, Ali Roth, Valentin Lampert, Christoph H. Machine Learning We study continual mean estimation, where data vectors arrive sequentially and the goal is to maintain accurate estimates of the running mean. We address this problem under user-level differential privacy, which protects each user's entire dataset even when they contribute multiple data points. Previous work on this problem has focused on pure differential privacy. While important, this approach limits applicability, as it leads to overly noisy estimates. In contrast, we analyze the problem under approximate differential privacy, adopting recent advances in the Matrix Factorization mechanism. We introduce a novel mean estimation specific factorization, which is both efficient and accurate, achieving asymptotically lower mean-squared error bounds in continual mean estimation under user-level differential privacy. |
| title | Matrix Factorization for Practical Continual Mean Estimation Under User-Level Differential Privacy |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.22320 |