Towards Trustworthy Collaborative Learning
Rachid Guerraoui, Professor, Department of Computer Science, EPFL
COM1 Level 2
Video Conference Room, COM1-02-13
Abstract:
With the rise of privacy concerns in training models, distributed approaches such as federated learning or decentralized learning (DL) systems are becoming increasingly attractive. Such systems enable nodes to collaboratively train a global ML model without sharing their private datasets with other entities. model updates being shared in such systems can be prone to privacy breaches and faulty nodes can significantly hamper the learning. This talk will present techniques to tolerate honest but curious nodes as well as Byzantine players.
Bio:
Anne-Marie Kermarrec is Professor of computer science at EPFL. Previously, she was the CEO of the startup Mediego, which she founded in April 2015.
From 2004 to 2015, she was a research director at Inria in France.
Anne-Marie Kermarrec earned a PhD from the University of Rennes (France) and has worked at the Vrije Universiteit Amsterdam (Netherlands) as well as at Microsoft Research in Cambridge (United Kingdom). She was elected to the Academia Europaea in 2013, to the French Academy of science in 2024, and named an ACM Fellow in 2016.
Bio:
Rachid Guerraoui is professor in computer science at EPFL where he leads the laboratory of Distributed Computing. He earned a PhD from the University of Orsay (France) and worked in the past with Ecole des Mines de Paris, CEA Saclay, HP Labs in Palo Alto and MIT. He has been elected ACM fellow and professor of the College de France. He was awarded a Senior ERC Grant, a Google Focused Award and the Dahl Nygaard Award. He was also elected the best professor in computer science by EPFL in 2024.