CS-E4740 - Federated Learning

course offered during spring 2024 at Aalto University

You can formally enrol this course as

• university (of applied science) student in Finland (contact your study administrator for details)
• adult learner in Finland via Non-Degree Studies at Aalto University
• student at Chalmers, KTH (Sweden), NTNU (Norway), DTU (Denmark) via Registration for Nordic Tech Five

Anybody interested in following the course (without formal enrolment) Subscribe to the course mailing list

Lectures *** What's New? *** Assignments *** FL Project

Abstract

Federated learning (FL) enables decentralized training of machine learning models, eliminating the need to collect local datasets at a central location. This course teaches the application of linear algebra and calculus to analyze and design FL systems, addressing real-world applications such as weather prediction and healthcare. You will learn to formulate FL applications as optimization problems and solve them with distributed algorithms. Students have the option to extend the course to 10 credits by completing a student project. This student project might be used to pilot ideas (e.g., for your doctoral research) and get peer feedback on them.

To get a more concrete idea of what to expect, have a look at the draft for the lecture notes.

References

[1] A. Jung, "Machine Learning. The Basics," Springer, Singapore, 2022. available via Aalto library here. preprint.

Additional Material

• Federated learning: Basics and application to the mobile keyboard (ML Tech Talks)
• Network Flows in Federated Learning (slides for an IEEE Seminar Talk)