Privacy in Federated and Collaborative Learning - Part I

Learn fundamental concepts of privacy in machine learning through this comprehensive tutorial delivered by Adam Smith from Boston University and Lydia Zakynthinou from Johns Hopkins University. Explore the theoretical foundations and practical implications of privacy-preserving techniques in computational systems, with particular emphasis on how privacy considerations impact federated and collaborative learning environments. Delve into key privacy models, understand the trade-offs between privacy and utility, and examine the mathematical frameworks that underpin modern privacy-preserving algorithms. Gain insights into differential privacy, its variants, and other privacy-preserving mechanisms that are essential for protecting sensitive data in distributed learning scenarios. Discover how privacy constraints shape algorithm design and performance in collaborative machine learning settings, preparing you for advanced topics in privacy-aware computational systems.