DIFFERENTIALLY PRIVATE LEARNING OF GEOMETRIC CONCEPTS

Research output: Contribution to journalArticlepeer-review

Abstract

We present efficient differentially private algorithms for learning unions of polygons in the plane (which are not necessarily convex). Our algorithms are (α , β)-probably approximately correct and (ϵ , δ )-differentially private using a sample of size O ( 1 α) , where the domain is [d] × [d] and k is the number of edges in the union of polygons. Our algorithms are obtained by designing a private variant of the classical (nonprivate) learner for conjunctions using the greedy algorithm for set cover.

Original languageEnglish
Pages (from-to)952-974
Number of pages23
JournalSIAM Journal on Optimization
Volume32
Issue number3
DOIs
StatePublished - 2022

Funding

FundersFunder number
Knut och Alice Wallenbergs Stiftelse
Vetenskapsrådet2018-05973

    Keywords

    • PAC learning
    • differential privacy
    • polygons

    Fingerprint

    Dive into the research topics of 'DIFFERENTIALLY PRIVATE LEARNING OF GEOMETRIC CONCEPTS'. Together they form a unique fingerprint.

    Cite this