Salvato in:
Dettagli Bibliografici
Autori principali: Fitzsimons, Jack, Honaker, James, Shoemate, Michael, Singhal, Vikrant
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2408.10438
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912256342622208
author Fitzsimons, Jack
Honaker, James
Shoemate, Michael
Singhal, Vikrant
author_facet Fitzsimons, Jack
Honaker, James
Shoemate, Michael
Singhal, Vikrant
contents We show that the most well-known and fundamental building blocks of DP implementations -- sum, mean, count (and many other linear queries) -- can be released with substantially reduced noise for the same privacy guarantee. We achieve this by projecting individual data with worst-case sensitivity $R$ onto a simplex where all data now has a constant norm $R$. In this simplex, additional ``free'' queries can be run that are already covered by the privacy-loss of the original budgeted query, and which algebraically give additional estimates of counts or sums.
format Preprint
id arxiv_https___arxiv_org_abs_2408_10438
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Private Means and the Curious Incident of the Free Lunch
Fitzsimons, Jack
Honaker, James
Shoemate, Michael
Singhal, Vikrant
Cryptography and Security
We show that the most well-known and fundamental building blocks of DP implementations -- sum, mean, count (and many other linear queries) -- can be released with substantially reduced noise for the same privacy guarantee. We achieve this by projecting individual data with worst-case sensitivity $R$ onto a simplex where all data now has a constant norm $R$. In this simplex, additional ``free'' queries can be run that are already covered by the privacy-loss of the original budgeted query, and which algebraically give additional estimates of counts or sums.
title Private Means and the Curious Incident of the Free Lunch
topic Cryptography and Security
url https://arxiv.org/abs/2408.10438