Improved Differentially Private Euclidean Distance Approximation
March 22, 2022 ยท Declared Dead ยท ๐ ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Nina Mesing Stausholm
arXiv ID
2203.11561
Category
cs.DS: Data Structures & Algorithms
Citations
13
Venue
ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems
Last Checked
3 months ago
Abstract
This work shows how to privately and more accurately estimate Euclidean distance between pairs of vectors. Input vectors $x$ and $y$ are mapped to differentially private sketches $x'$ and $y'$, from which one can estimate the distance between $x$ and $y$. Our estimator relies on the Sparser Johnson-Lindenstrauss constructions by Kane \& Nelson (Journal of the ACM 2014), which for any $0<ฮฑ,ฮฒ<1/2$ have optimal output dimension $k=ฮ(ฮฑ^{-2}\log(1/ฮฒ))$ and sparsity $s=O(ฮฑ^{-1}\log(1/ฮฒ))$. We combine the constructions of Kane \& Nelson with either the Laplace or the Gaussian mechanism from the differential privacy literature, depending on the privacy parameters $\varepsilon$ and $ฮด$. We also suggest a differentially private version of Fast Johnson-Lindenstrauss Transform (FJLT) by Ailon \& Chazelle (SIAM Journal of Computing 2009) which offers a tradeoff in speed for variance for certain parameters. We answer an open question by Kenthapadi et al.~(Journal of Privacy and Confidentiality 2013) by analyzing the privacy and utility guarantees of an estimator for Euclidean distance, relying on Laplacian rather than Gaussian noise. We prove that the Laplace mechanism yields lower variance than the Gaussian mechanism whenever $ฮด<ฮฒ^{O(1/ฮฑ)}$. Thus, our work poses an improvement over the work of Kenthapadi et al.~by giving a more efficient estimator with lower variance for sufficiently small $ฮด$. Our sketch also achieves \emph{pure} differential privacy as a neat side-effect of the Laplace mechanism rather than the \emph{approximate} differential privacy guarantee of the Gaussian mechanism, which may not be sufficiently strong for some settings.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
๐ Similar Papers
In the same crypt โ Data Structures & Algorithms
R.I.P.
๐ป
Ghosted
R.I.P.
๐ป
Ghosted
Relief-Based Feature Selection: Introduction and Review
R.I.P.
๐ป
Ghosted
Route Planning in Transportation Networks
R.I.P.
๐ป
Ghosted
Near-linear time approximation algorithms for optimal transport via Sinkhorn iteration
R.I.P.
๐ป
Ghosted
Hierarchical Clustering: Objective Functions and Algorithms
R.I.P.
๐ป
Ghosted
Graph Isomorphism in Quasipolynomial Time
Died the same way โ ๐ป Ghosted
R.I.P.
๐ป
Ghosted
Language Models are Few-Shot Learners
R.I.P.
๐ป
Ghosted
PyTorch: An Imperative Style, High-Performance Deep Learning Library
R.I.P.
๐ป
Ghosted
XGBoost: A Scalable Tree Boosting System
R.I.P.
๐ป
Ghosted