Federated Learning: Collaborative Machine Learning without Centralized Training Data

Brendan McMahan, Daniel Ramage, Google Research Blog, Apr 10, 2017
Commentary by Stephen Downes
files/images/FederatedLearning_FinalFiles_Flow_Chart1.png

One of the problems with learning analytics and analytics in general is that it requires a lot of data. This means you have to watch what a lot of people are doing, which has ethical and privacy implications. The federated analytics model described here attempts to address these issues. "Your device downloads the current model, improves it by learning from data on your phone, and then summarizes the changes as a small focused update. Only this update to the model is sent to the cloud." Of course, you have to trust that your device is actually doing this.

Views: 0 today, 907 total (since January 1, 2017).[Direct Link]
Creative Commons License. gRSShopper

Copyright 2015 Stephen Downes ~ Contact: stephen@downes.ca
This page generated by gRSShopper.
Last Updated: Aug 23, 2017 02:15 a.m.