Speaker: Keith Bonawitz (Google Research)

Title: Federated Learning: Privacy-Preserving Collaborative Machine Learning without Centralized Training Data

Abstract: Federated Learning enables mobile devices to collaboratively learn a shared inference model while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud. In this talk, I will discuss (1) the challenges that distinguish Federated Learning from other distributed learning scenarios, (2) the algorithms and systems we've created to address those challenges, (3) extensions to those algorithms to minimize communication costs, and (4) the novel cryptographic and randomized privacy-preserving technologies we have developed for Federated Learning.