Data Privacy In Toto: Quantification, Synthesis, Learning

Date
Mar 29, 2016, 11:00 am12:00 pm
Location
Engineering Quadrangle, Room B205

Speaker

Details

Event Description

Abstract:  In this talk, I will use location privacy as a case study in how we can address the hardest privacy problems in a rigorous and comprehensive fashion. Location data is widely collected by service providers and mobile apps, yet is very difficult to protect: it is sparse, strongly identifiable, highly sensitive, vulnerable to inadvertent disclosure, and easily correlated with widely available auxiliary information. I will present my approach to quantifying privacy, which has been broadly adopted by the research community, as well as a systematic methodology for achieving optimal privacy while preserving data utility. I will then discuss my ongoing research at the junction of privacy, data science, and machine learning in two emerging scenarios: generating privacy-preserving synthetic data and building accurate deep-learning models that respect privacy of the training data.
 
Biography:  Reza Shokri is a Postdoctoral Researcher at Cornell University. His research focuses on quantitative analysis of privacy, as well as design of privacy-preserving systems for a variety of applications, from location-based services and recommender systems to web search and machine learning. His work on quantifying location privacy was recognized as a runner-up for the annual Award for Outstanding Research in Privacy Enhancing Technologies (PET Award). Recently, he has focused on privacy-preserving generative models, and privacy-preserving deep learning.  He received his PhD from EPFL, and spent one year as a postdoctoral researcher at ETH Zurich.