This guide provides a reference framework for trustworthy Federated Machine Learning. The document provides guidance with respect to provable security for data and models, optimized model utility, controllable communication and computational complexity, explainable decision making and supervised processes. It describes three main aspects:
- Principles for trustworthy Federated Machine Learning
- Requirements for different roles in trustworthy Federated Machine Learning
- Techniques to realize trustworthy Federated Machine Learning
The purpose of this guide is to provide credible, practical and controllable solution guidance for Federated Machine Learning and other privacy computing applications.
