An Overview of Differential Privacy, Membership Inference Attacks, and Federated Learning
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact James Allingham.
Zoom link available upon request (it is sent out on our mailing list, eng-mlg-rcc [at] lists.cam.ac.uk). Sign up to our mailing list for easier reminders.
This tutorial will cover the basics of differential privacy (DP) including the Gaussian mechanism, training networks with DP-SGD, and a look at various state-of-the-art approaches. We then describe the ideas behind membership inference attacks and show how they can be used to audit differentially private systems. Finally, we give an overview of federated learning and explain how it can be made to be differentially private. If there is time remaining, we will present a case study on differentially private speech.
Required Reading: None.
This talk is part of the Machine Learning Reading Group @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|