Differentially Private Learning of Geometric Concepts

02/13/2019

∙

We present differentially private efficient algorithms for learning union of polygons in the plane (which are not necessarily convex). Our algorithms achieve (α,β)-PAC learning and (ϵ,δ)-differential privacy using a sample of size Õ(1/αϵk d), where the domain is [d]×[d] and k is the number of edges in the union of polygons.

READ FULL TEXT

Differentially Private Learning of Geometric Concepts

Differentially Private Policy Evaluation

Differentially Private Identity and Closeness Testing of Discrete Distributions

Private Learning Implies Online Learning: An Efficient Reduction

Faster Differentially Private Samplers via Rényi Divergence Analysis of Discretized Langevin MCMC

Differentially Private Supervised Manifold Learning with Applications like Private Image Retrieval

Differentially Private Shapley Values for Data Evaluation

Private Approximations of a Convex Hull in Low Dimensions

Differentially Private Learning of Geometric Concepts

Related Research

Differentially Private Policy Evaluation

Differentially Private Identity and Closeness Testing of Discrete Distributions

Private Learning Implies Online Learning: An Efficient Reduction

Faster Differentially Private Samplers via Rényi Divergence Analysis of Discretized Langevin MCMC

Differentially Private Supervised Manifold Learning with Applications like Private Image Retrieval

Differentially Private Shapley Values for Data Evaluation

Private Approximations of a Convex Hull in Low Dimensions