INSPECTRE: Privately Estimating the Unseen

Wednesday, April 11, 2018 - 4:30pm to 5:00pm

Event Calendar Category

LIDS & Stats Tea

Speaker Name

Gautam Kamath

Affiliation

CSAIL

Building and Room Number

LIDS Lounge

Abstract

Gautam Kamath will discuss some his recent work on estimating distribution properties, with the additional constraint of guaranteeing the privacy of the sample. Some properties of interest include entropy, support size, and support coverage, which have enjoyed a significant interdisciplinary study in the computer science, information theory, and statistics communities. In all these cases, privacy comes at almost no additional cost over the standard estimation problem.

Joint work with Jayadev Acharya, Ziteng Sun, and Huanyu Zhang.