Lecturers
Abstract for the course
Differential Privacy offers ways to answer statistical queries about sensitive data while providing strong provable privacy guarantees ensuring that the presence or absence of a single individual in the data has a negligible statistical effect on the query's result. Differential privacy is becoming a gold standard in data privacy and it is now used by statistical agencies like the US Census Bureau and companies like Apple, Google, and Uber. This course will focus on fundamental results about the theory and practice of differential privacy and how to use it in concrete applications. In particular, the course will introduce students to some fundamental mechanisms for building differentially private queries. The course will also present a novel framework for programming such queries and how to use them to design privacy-preserving applications that take into account not only the privacy of individuals but also the accuracy of obtained statistical results. This tool has been developed by the speakers and their collaborators.
Prerequisites
The only prerequisite for students is to understand basic probability theory and basic skills on functional programming. The rest of the course is self-contained.
Acknowledgments
The development of this course is partly supported the Swedish research agency VR, the SSF Swedish Foundation for Strategic Research (SSF) under the project Octopi (Ref. RIT17-0023) and WebSec (Ref. RIT17-0011).