differentially private median selection

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering

Final Year Thesis Oral Defense

Title: "differentially private median selection"

By

HUANG Meng

Abstract:

Differential privacy is the standard way for defining and protecting 
privacy in a mathematically rigorous way. In simple terms, it means that 
privacy is protected if the result of a query cannot confidently imply any 
single individual's data in the database. It is used by big companies like 
Apple and Microsoft to protect customers' privacy when conducting 
data-related studies. In this thesis, we examine a specific problem in 
differential privacy, that is, median calculation, a frequently used SQL 
operator. We have implemented, analyzed, evaluated and compared several 
existing methods for differentially private median calculation. We 
collected statistics on their performance on simulated and real-world 
data.


Date            : 5 May 2022 (Thursday)

Time            : 09:00-09:40

Zoom Link:
https://hkust.zoom.us/j/93306848754?pwd=UmEvZWFXSnFBMDdyTHhyVHpQQkx0UT09

Meeting ID      : 933 0684 8754

Passcode        : 571821

Advisor         : Prof. YI Ke

2nd Reader      : Prof. GOLIN Mordecai J.