Federated Learning: Challenges, Methods, and Future Directions

Devices communicate with a central server periodically to learn a global model. Federated learning helps preserve user privacy and reduce strain on the network by keeping data localized. How does it differ from traditional large-scale machine learning, distributed optimization, and privacy-preserving data analysis? What do we understand currently about federated learning, and what problems are left to explore? In this post, we briefly answer these questions, and describe ongoing work in federated learning at CMU.