Many applications in modern computing (AI, ML and data science) rely on the data from many individuals. However, this data is often sensitive, and needs to be handled under strong guarantees of privacy and security.
I'll discuss a range of approaches that can be used at scale to gather statistics and build models of data while ensuring the strong guarantees are met. These combine techniques from differential privacy, federated learning, and cryptography.
Link to join virtually: https://cam-ac-uk.zoom.us/j/81322468305
This talk is being recorded, accessible to members of the University.