Private Center Points and Learning of Halfspaces
Beimel, Amos, Moran, Shay, Nissim, Kobbi, Stemmer, Uri
–arXiv.org Artificial Intelligence
We present a private learner for halfspaces over an arbitrary finite domain $X\subset \mathbb{R}^d$ with sample complexity $mathrm{poly}(d,2^{\log^*|X|})$. The building block for this learner is a differentially private algorithm for locating an approximate center point of $m>\mathrm{poly}(d,2^{\log^*|X|})$ points -- a high dimensional generalization of the median function. Our construction establishes a relationship between these two problems that is reminiscent of the relation between the median and learning one-dimensional thresholds [Bun et al.\ FOCS '15]. This relationship suggests that the problem of privately locating a center point may have further applications in the design of differentially private algorithms. We also provide a lower bound on the sample complexity for privately finding a point in the convex hull. For approximate differential privacy, we show a lower bound of $m=\Omega(d+\log^*|X|)$, whereas for pure differential privacy $m=\Omega(d\log|X|)$.
arXiv.org Artificial Intelligence
Feb-27-2019
- Country:
- Asia > Middle East
- Israel (0.14)
- Europe > Sweden (0.14)
- North America
- Canada (0.14)
- United States (0.14)
- Asia > Middle East
- Genre:
- Research Report (0.64)
- Industry:
- Information Technology > Security & Privacy (0.93)
- Technology: