Dancing in the Dark: Private Multi-Party Machine Learning in an Untrusted Setting

Friday December 15th, 12-1PM @ BA5205

Speaker: Clement Fung

Title:
Dancing in the Dark: Private Multi-Party Machine Learning in an Untrusted Setting

Abstract:
The problem of machine learning (ML) over distributed data sources arises in a variety of domains. Unfortunately, today’s distributed ML systems use an unsophisticated threat model:
data sources must trust a central ML process. We propose a brokered learning abstraction that provides data sources with provable privacy guarantees while allowing them to contribute data towards a globally-learned model in an untrusted setting. We realize this abstraction by building on the state of the art in multi-party distributed ML and differential privacy methods to construct TorMentor, a system that is deployed as a hidden Tor service.

Bio:
Clement Fung is a second year master’s student in Computer Science at UBC, supervised by Prof. Ivan Beschastnikh. Originally from Toronto, he completed his undergraduate at the University of Waterloo in 2016. His interests are in privacy-preserving machine learning and distributed systems.