English

Private Selection from Private Candidates

Data Structures and Algorithms 2018-11-21 v1 Cryptography and Security Machine Learning Machine Learning

Abstract

Differentially Private algorithms often need to select the best amongst many candidate options. Classical works on this selection problem require that the candidates' goodness, measured as a real-valued score function, does not change by much when one person's data changes. In many applications such as hyperparameter optimization, this stability assumption is much too strong. In this work, we consider the selection problem under a much weaker stability assumption on the candidates, namely that the score functions are differentially private. Under this assumption, we present algorithms that are near-optimal along the three relevant dimensions: privacy, utility and computational efficiency. Our result can be seen as a generalization of the exponential mechanism and its existing generalizations. We also develop an online version of our algorithm, that can be seen as a generalization of the sparse vector technique to this weaker stability assumption. We show how our results imply better algorithms for hyperparameter selection in differentially private machine learning, as well as for adaptive data analysis.

Keywords

Cite

@article{arxiv.1811.07971,
  title  = {Private Selection from Private Candidates},
  author = {Jingcheng Liu and Kunal Talwar},
  journal= {arXiv preprint arXiv:1811.07971},
  year   = {2018}
}

Comments

38 pages

R2 v1 2026-06-23T05:21:24.783Z