LFU:online

Getting familiar with the main theoretical and practical concepts of Machine Learning and its prospective use for the study of classical and quantum physics.

The course will cover key concepts of machine learning (ML) from a theoretical perspective while maintaining a strong focus on practical applications. In particular, it will examine how a variety of ML methods can be employed in both classical and quantum physics.

To that end, we will first review the fundamentals of machine learning, from the different tasks that ML can address to the training of simple neural networks. We will then turn to more advanced topics—ranging from diverse ML architectures to the training of deep-learning models—and conclude with typical ML applications in physics, such as analyzing and characterizing physical systems, controlling experiments or discovering new physical concepts.

Throughout, the course will emphasize hands-on implementation in Python, using established ML libraries such as PyTorch and fastai.

Evaluation based on class participation, coding homeworks and a final project.

Lehrveranstaltungsprüfung gemäß § 7 Satzungsteil, Studienrechtliche Bestimmungen.

The primary reference will be the book "Machine Learning in Quantum Sciences" (link; freely available online here). 

Basic knowledge of Python is highly recommended.