AI Paper Atlas

Research map / Federated learning and privacy / Differential privacy

Oneshot Differentially Private Top-k Selection

Gang Qiao, Weijie Su, Li Zhang

ICML 2021 · Federated learning and privacy › Differential privacy · approach: Privacy accounting and mechanisms · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Select all k items from one noisy ranking rather than running Report Noisy Max k timesOneshot Laplace · Top-k selectionnumber of selection passes: k times -> onceImproved
Use one-shot selection with Laplace noise calibrated for approximate privacyTop-k selection · Top-k selectionnoise scale for (ε,δ)-differential privacy: \widetilde{O}(\sqrt{k}/\eps)Improved
Apply the one-shot mechanism to stationary-distribution scores from pairwise comparisonsOneshot differentially private spectral method · Private pairwise comparison top-knoise scale for (ε,δ)-differential privacy: \lambda=\frac{8s\sqrt{k\log(m/\delta)}}{\e}Improved
Use the one-shot mechanism when adjacent inputs have a sufficiently large minimum gapOneshot Laplace · Top-k selection with \Delta \ge 20 \lambda and m \le 8\times 10^6probability of returning the true top-k set: p(\Delta)>0.99Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy