A Survey on Differential Privacy

PhD Qualifying Examination


Title: "A Survey on Differential Privacy"

by

Mr. Ziyue HUANG


Abstract:

The past few years have seen major advance and success in the area of
machine learning and data analysis, and there is an increasing trend for
collecting and analyzing personal data which might be sensitive, such as
disease, salary and financial information. To protect the privacy of the
involved individuals, while still allowing a data analyst to derive useful
statistics, various definitions and frameworks are pro- posed. In this
survey, we will focus on differential privacy introduced by Dwork et al.
in 2006, where a trusted curator holds the entire data set of sensitive
information and it is guaranteed that the published statistics is
not affected (by much) by any individual. We firstly introduce the formal
definition of differential privacy and its basic properties. Then we
describe several useful and fundamental mechanisms satisfying differential
privacy and their real world applications. Finally we consider an
extension of differential privacy to the local model without the trust
assumption on a data curator.


Date:                   Thursday, 20 February 2020

Time:                   3:00pm - 5:00pm

Zoom Meeting:           https://hkust.zoom.us/j/168991258

Committee Members:      Prof. Ke Yi (Supervisor)
                        Prof. Cunsheng Ding (Chairperson)
                        Dr. Dimitris Papadopoulos
                        Dr. Raymond Wong


**** ALL are Welcome ****