English

Continual Release Moment Estimation with Differential Privacy

Machine Learning 2025-06-05 v2 Machine Learning

Abstract

We propose Joint Moment Estimation (JME), a method for continually and privately estimating both the first and second moments of data with reduced noise compared to naive approaches. JME uses the matrix mechanism and a joint sensitivity analysis to allow the second moment estimation with no additional privacy cost, thereby improving accuracy while maintaining privacy. We demonstrate JME's effectiveness in two applications: estimating the running mean and covariance matrix for Gaussian density estimation, and model training with DP-Adam on CIFAR-10.

Keywords

Cite

@article{arxiv.2502.06597,
  title  = {Continual Release Moment Estimation with Differential Privacy},
  author = {Nikita P. Kalinin and Jalaj Upadhyay and Christoph H. Lampert},
  journal= {arXiv preprint arXiv:2502.06597},
  year   = {2025}
}
R2 v1 2026-06-28T21:38:46.318Z