CRDBFeb 2, 2022

Exact Privacy Analysis of the Gaussian Sparse Histogram Mechanism

arXiv:2202.01100v19 citations
Originality Synthesis-oriented
AI Analysis

This work provides more precise privacy analysis for sparse histogram methods, which is incremental but important for applications like large group-by queries.

The authors tackled the problem of quantifying the exact differential privacy guarantees of the Gaussian sparse histogram mechanism, finding that prior overestimates led to looser privacy bounds.

Sparse histogram methods can be useful for returning differentially private counts of items in large or infinite histograms, large group-by queries, and more generally, releasing a set of statistics with sufficient item counts. We consider the Gaussian version of the sparse histogram mechanism and study the exact $ε,δ$ differential privacy guarantees satisfied by this mechanism. We compare these exact $ε,δ$ parameters to the simpler overestimates used in prior work to quantify the impact of their looser privacy bounds.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes