Private Distribution Learning with Public Data: The View from Sample Compression

We study the problem of private distribution learning with access to public data. In this setup, which we refer to as public-private learning, the learner is given public and private samples drawn from an unknown distribution belonging to a class , with the goal of outputting an estimate of while adhering to privacy constraints (here, pure differential privacy) only with respect to the private samples. We show that the public-private learnability of a class is connected to the existence of a sample compression scheme for , as well as to an intermediate notion we refer to as list learning. Leveraging this connection: (1) approximately recovers previous results on Gaussians over ; and (2) leads to new ones, including sample complexity upper bounds for arbitrary -mixtures of Gaussians over , results for agnostic and distribution-shift resistant learners, as well as closure properties for public-private learnability under taking mixtures and products of distributions. Finally, via the connection to list learning, we show that for Gaussians in , at least public samples are necessary for private learnability, which is close to the known upper bound of public samples.
View on arXiv