Watch a 37-minute research talk from Yi Yu of the University of Warwick exploring distributed data scenarios with varying differential privacy constraints across servers and users. Delve into the challenges of transmitting private information between servers/users and examine fundamental limits in classical statistical inference problems within this context. Learn about domain adaptation considerations when dealing with user-level privacy and federated learning environments where data remains distributed across multiple locations. Gain insights into privacy-preserving data analysis techniques that maintain security while enabling meaningful statistical analysis across distributed systems.
User-Level and Federated Local Differential Privacy in Statistical Inference
Overview
Syllabus
User-level and federated local differential privacy
Taught by
Simons Institute
Related Courses
-
Federated Learning with Formal User-Level Differential Privacy Guarantees
-
Privacy-Aware Compression for Federated Learning
-
Federated Heavy Hitters Discovery with Differential Privacy
-
Preserving Data Privacy in Federated Learning - Xiaokui Xiao
4.0 -
Private Multi-Task Learning - Formulation and Applications to Federated Learning
-
Federated Learning for Nonparametric Function Estimation - Framework and Optimality
