![Modern Paradigms in Generalization logo](/sites/default/files/styles/workshop_banner_sm_1x/public/2024-04/Modern%20Paradigms%20in%20Generalization%20logo_0.jpg?h=0100a3be&itok=Eiim-c68)
Abstract
This talk aims to discuss the case when data are distributed and different differential privacy constraints are imposed on each of the servers/users. Across servers/users, only private information is allowed to be transmitted. Different classical statistical inference problems will be discussed, accompanied with their fundamental limits.