Federated Learning
摘要
This chapter gives a quick introduction to distributed machine learning, and then moves onto federated learning, i.e., large-scale distributed machine learning powered mainly by deep learning. Besides the algorithms for the new distributed paradigm, we will also look at the data privacy issue, which arises from the network settings we now adopt and is, a little surprisingly, not an easy fix in distributed deep learning in the open. Finally, readers are pointed to some useful online resources and readings for further explorations.