English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this work, we study high-dimensional mean estimation under user-leveldifferential privacy, and design an $(\varepsilon,\delta)$-differentiallyprivate mechanism using as few users as possible. In particular, we provide anearly optimal trade-off between the number of users and the number of samplesper user required for private mean estimation, even when the number of users isas low as $O(\frac{1}{\varepsilon}\log\frac{1}{\delta})$. Interestingly, thisbound on the number of \emph{users} is independent of the dimension (though thenumber of \emph{samples per user} is allowed to depend polynomially on thedimension), unlike the previous work that requires the number of users todepend polynomially on the dimension. This resolves a problem first proposed byAmin et al. Moreover, our mechanism is robust against corruptions in up to$49\%$ of the users. Finally, our results also apply to optimal algorithms forprivately learning discrete distributions with few users, answering a questionof Liu et al., and a broader range of problems such as stochastic convexoptimization and a variant of stochastic gradient descent via a reduction todifferentially private mean estimation.

Tight and Robust Private Mean Estimation with Few Users