Differentially Private Query Release Through Adaptive Projection

We propose, implement, and evaluate a new algorithm for releasing answers to very large numbers of statistical queries like k-way marginals, subject to differential privacy. Our algorithm makes adaptive use of a continuous relaxation of the Projection Mechanism, which answers queries on the private dataset using simple perturbation, and then attempts to find the synthetic dataset that most closely matches the noisy answers. We use a continuous relaxation of the synthetic dataset domain which makes the projection loss differentiable, and allows us to use efficient ML optimization techniques and tooling. Rather than answering all queries up front, we make judicious use of our privacy budget by iteratively finding queries for which our (relaxed) synthetic data has high error, and then repeating the projection. Randomized rounding allows us to obtain synthetic data in the original schema. We perform experimental evaluations across a range of parameters and datasets, and find that our method outperforms existing algorithms on large query classes.