Professor of Computer Science Vitaly Shmatikov‘s work on privacy in machine learning recently received the Caspar Bowden Award for Outstanding Research in Privacy Enhancing Technologies.
This work was done by postdoctoral fellows Reza Shokri and Marco Stronati, PhD student Congzheng Song, and Professor Shmatikov.
In a recent blog post on Freedom to Tinker, Shmatikov discussed the findings of the study:
We uncovered multiple privacy and integrity problems in today’s ML pipelines, especially (1) online services such as Amazon ML and Google Prediction API that create ML models on demand for non-expert users, and (2) federated learning, aka collaborative learning, that lets multiple users create a joint ML model while keeping their data private (imagine millions of smartphones jointly training a predictive keyboard on users’ typed messages).”
Professor Shmatikov’s research group also received this award in 2008 and 2014.