BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Noise-Aware Differentially Private Synthetic Data - Antti Honkela\
 , University of Helsinki
DTSTART:20220628T100000Z
DTEND:20220628T110000Z
UID:TALK176027@talks.cam.ac.uk
CONTACT:Dr R.E. Turner
DESCRIPTION:Synthetic data generated under differential privacy (DP) promi
 ses to significantly simplify analysis of sensitive personal data. Existin
 g work has shown that simply analysing DP synthetic data as if it were rea
 l does not produce valid inferences of population-level quantities\, leadi
 ng to too narrow confidence intervals and thereby risking false discoverie
 s. We propose using multiple imputation techniques to avoid these problems
 . This requires simulating multiple synthetic data sets from the Bayesian 
 posterior predictive distribution over data sets. We propose a novel noise
 -aware Bayesian DP synthetic data generation mechanism for discrete data t
 hat enables generating such a distribution of data sets. Our experiments d
 emonstrate that the method is able to produce accurate confidence interval
 s from DP synthetic data.
LOCATION:Hybrid\, CBL Seminar room\, Department of Engineering\, and Zoom 
 https://eng-cam.zoom.us/j/89002493651?pwd=B_2gKl7va_h0CQ9yoMPSbn2ifYLGi4.1
END:VEVENT
END:VCALENDAR