ELLIS Summer School co-organised by Unit Munich at TU Dortmund University

In collaboration with ELLIS Unit Munich, the summer school covers foundational and recent topics in AutoML and will last 5 days. This year, we will add a special focus on agentic systems for solving machine learning and data science workflows. The schedule covers three different lecture types and interactive elements: introductory lectures, advanced lectures, and hands-on sessions. The introductory lectures bring the participants up to speed on basic topics in AutoML, covering fundamental concepts like global optimization, Bayesian optimization, and neural architecture search. The advanced lectures cover current research directions AutoML with and for large pre-trained models. The hands-on sessions introduce students to practical AutoML tools and frameworks that enable them to use AutoML techniques or conduct novel research on AutoML, for example, by giving access to state-of-the-art Bayesian optimization tools (e.g., SMAC3) or AutoML systems (e.g., AutoGluon).

Contact: Matthias Feurer & Katharina Eggensperger

This is part of the 2026 ELLIS Winter & Summer School Program