TCS+ talk: Wednesday, October 23, Thomas Steinke, Google DeepMind
Clement Canonne
Our next talk will take place this coming Wednesday, October 23th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 17:00 UTC). Thomas Steinke from Google DeepMind will speak about "The Discrete Gaussian for Differential Privacy" (abstract below).
Please sign up on the online form at https://sites.google.com/view/tcsplus/welcome/next-tcs-talk if you wish to join the talk as an individual or a group. Registration is /not/ required to attend the interactive talk, and the link will be posted on the website the day prior to the talk; however, by registering in the form, you will receive a reminder, along with the link. (The link to the recording will also be posted on our website afterwards.)
Hoping to see you all there,
The organizers
-------------------------------
Speaker: Thomas Steinke (Google DeepMind)
Title: The Discrete Gaussian for Differential Privacy
Abstract: A key tool for building differentially private systems is adding Gaussian noise to the output of a function evaluated on a sensitive dataset. Unfortunately, using a continuous distribution presents several practical challenges. First and foremost, finite computers cannot exactly represent samples from continuous distributions, and previous work has demonstrated that seemingly innocuous numerical errors can entirely destroy privacy. Moreover, when the underlying data is itself discrete (e.g., population counts), adding continuous noise makes the result less interpretable.
With these shortcomings in mind, we introduce and analyze the discrete Gaussian in the context of differential privacy. Specifically, we theoretically and experimentally show that adding discrete Gaussian noise provides essentially the same privacy and accuracy guarantees as the addition of continuous Gaussian noise. We also present an simple and efficient algorithm for exact sampling from this distribution. This demonstrates its applicability for privately answering counting queries, or more generally, low-sensitivity integer-valued queries.
Clement Canonne
The link for tomorrow's TCS+ talk has been posted: you will be able to join tomorrow, starting at 12:50pm ET:
https://berkeley.zoom.us/j/98954371813?pwd=V1hxN2Nrc2c5OEJFSWRqS29JeWM1dz09 (you will need a to be logged in on Zoom to join: a free account suffices)
Best,
-- Clément, on behalf of the TCS+ team
________________________________________
From: 'Clement Canonne' via TCS+ <tcsplus_...@googlegroups.com>
Sent: Saturday, 19 October 2024 8:26 AM
To: tcsplus_...@googlegroups.com
Subject: TCS+ talk: Wednesday, October 23, Thomas Steinke, Google DeepMind
Dear TCS+ followers,
Our next talk will take place this coming Wednesday, October 23th at 1:00 PM Eastern Time (10:00 AM Pacific Time, 19:00 Central European Time, 17:00 UTC). Thomas Steinke from Google DeepMind will speak about "The Discrete Gaussian for Differential Privacy" (abstract below).
The organizers
-------------------------------
Speaker: Thomas Steinke (Google DeepMind)
Title: The Discrete Gaussian for Differential Privacy
Abstract: A key tool for building differentially private systems is adding Gaussian noise to the output of a function evaluated on a sensitive dataset. Unfortunately, using a continuous distribution presents several practical challenges. First and foremost, finite computers cannot exactly represent samples from continuous distributions, and previous work has demonstrated that seemingly innocuous numerical errors can entirely destroy privacy. Moreover, when the underlying data is itself discrete (e.g., population counts), adding continuous noise makes the result less interpretable.
With these shortcomings in mind, we introduce and analyze the discrete Gaussian in the context of differential privacy. Specifically, we theoretically and experimentally show that adding discrete Gaussian noise provides essentially the same privacy and accuracy guarantees as the addition of continuous Gaussian noise. We also present an simple and efficient algorithm for exact sampling from this distribution. This demonstrates its applicability for privately answering counting queries, or more generally, low-sensitivity integer-valued queries.
You received this message because you are subscribed to the Google Groups "TCS+" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tcsplus_announ...@googlegroups.com.
To view this discussion on the web visit https://url.au.m.mimecastprotect.com/s/nDPYCK1DvKTr2RZ5EsvhWF5Iscm?domain=groups.google.com.
For more options, visit https://url.au.m.mimecastprotect.com/s/I_yECL7EwMfYRzlnjTPi8FynYs9?domain=groups.google.com.