Explore a cutting-edge secure aggregation protocol for federated learning in this 17-minute IEEE conference talk. Delve into ELSA (Efficient and Lightweight Secure Aggregation), a novel approach designed to address the challenges of malicious actors in federated learning environments. Learn how researchers from the University of California, Berkeley, Stanford University, and Devron Corporation have developed this protocol to enhance privacy and security in collaborative machine learning. Gain insights into the innovative techniques employed by ELSA to protect against potential threats while maintaining efficiency and scalability in federated learning systems.
ELSA - Secure Aggregation for Federated Learning with Malicious Actors
Overview
Syllabus
ELSA: Secure Aggregation for Federated Learning with Malicious Actors
Taught by
IEEE Symposium on Security and Privacy
Tags
Related Courses
-
Flamingo - Multi-Round Single-Server Secure Aggregation with Applications to Private Federated Learning
-
RoFL - Robustness of Secure Federated Learning
-
Federated Learning - Machine Learning on Decentralized Data
5.0 -
ADI - Adversarial Dominating Inputs in Vertical Federated Learning Systems
-
ACORN - Input Validation for Secure Aggregation
-
Comprehensive Privacy Analysis of Deep Learning
