Cray Speaker: Differential Privacy and the US Census

Colloquia
Cray Distinguished Speaker Series
May 4, 2020 - 11:15am to 12:15pm
Cynthia Dwork
Affiliation: 
Harvard
Location: 
Zoom Webinar: https://umn.zoom.us/j/98861668076
Host: 
Steven Wu
Abstract: Differential privacy is a mathematically rigorous definition of privacy tailored to statistical analysis of large datasets.  A driving scenario for the development of differential privacy was the US census: the people’s data, used to allocate the people’s resources, with a legal mandate for privacy.   Billions of dollars of federal spending, and, every decade, the allocation of representatives to states and the redrawing of congressional districts, depend on census data.  In 2020, for the first time, the confidentiality of individuals in the decennial census will be protected by differential privacy.
 
This talk will motivate the definition and use of differential privacy, reflect on the theory-meets-practice experiences of the decennial census, and highlight a few pressing  challenges in the field.
 
Bio: Cynthia Dwork, the Gordon McKay Professor of Computer Science at the John A. Paulson School of Engineering and Applied Sciences at Harvard, the Radcliffe Alumnae Professor at the Radcliffe Institute for Advanced Study,  Affiliated Faculty Member at Harvard Law School, and Distinguished Scientist at Microsoft Research, is renowned for spearheading the invention of differential privacy and the development of the field.  She has also made seminal contributions in distributed computing and cryptography.  Her current focus is on developing the theory of algorithmic fairness. Dwork is a member of the US National Academy of Sciences, the US National Academy of Engineering, and the American Philosophical Society, and is a Fellow of the American Academy of Arts and Sciences and of the ACM.