Seminars

View all Seminars  |  Download ICal for this event

Stable Estimators for Differentially Private Statistics

Series: Department Seminar

Speaker: Gavin Brown, Assistant Professor, Department of Computer Sciences, University of Wisconsin–Madison

Date/Time: Jan 07 11:00:00

Location: CSA Auditorium, (Room No. 104, Ground Floor)

Abstract:
We will discuss a line of work on techniques for stable statistical estimation, leading to efficient differentially private algorithms for mean estimation, covariance estimation, and linear regression. These algorithms are equivariant to affine transformations, leading to error guarantees that match those of the non-private empirical estimators under weak assumptions. The analysis proceeds by constructing a stabilizing wrapper around a greedy outlier-removal process.


Based on work with Sam Hopkins, Adam Smith, Jon Hayase, Xiyang Liu, Weihao Kong, Sewoong Oh, and Juan Perdomo.

Speaker Bio:
Gavin Brown is an Assistant Professor in the Department of Computer Sciences at the University of Wisconsin–Madison, where he leads a research group focused on machine learning and privacy. Before coming to Madison, he was a postdoctoral scholar at the University of Washington with Sewoong Oh. He completed his PhD under the supervision of Adam Smith at Boston University. Personal website: https://pages.cs.wisc.edu/~grbrown5/

Host Faculty: Prof. Vinod Ganapathy