Colloquium – Feb 27, 2024
with Dr. Simon Hirlaender

ScaDS.AI announces and welcomes you to join its public colloquium session on Tuesday, February 27, 2024 at 14:00-15:30 CET. The colloquium takes place at the large seminar room at ScaDS.AI Leipzig (details below) and in parallel online (link to Zoom session).

Dr. Nico Scherf, MPI of Cognitive and Brain Sciences (Leipzig, Germany), invites

Simon Hirlaender (PLUS, Salzburg, Austria)

Simon Hirlaender


Head of the SARL Team, IDA Lab,
Paris Lodron University of Salzburg (PLUS), Austria

Learn how to make good decisions under uncertainty:
Reinforcement learning and the real world

The presentation introduces reinforcement learning (RL) as a key machine learning technique for tackling sequential decision problems and positions it as key to unlocking the potential for general artificial intelligence. This field, currently at the forefront of academic and practical research, is explored through the lens of its application in real-world contexts, with the control of accelerators serving as a prime example. The discussion moves from deep meta-RL to safe model-based RL strategies and how RL can be tailored to the requirements of complex, variable accelerator environments such as at CERN. Critical issues such as safety, sample utilization efficiency and system adaptability will also be addressed. The seminar emphasizes the need to further develop RL methods to cope with the difficulties and uncertainties of using these technologies in real-world scenarios and highlights the need for innovative strategies to overcome these obstacles.

About Simon Hirlaender

Location

ScaDS.AI Leipzig
Löhrs Carré, Humboldtstrasse 25, 04105 Leipzig
3rd floor, large seminar room (A 03.07 „Zwenkauer See“)

In parallel online via Zoom (for the link see above, during the session only)

Check out our event calendar for more information about our upcoming events!

TU
Universität
Max
Leibnitz-Institut
Helmholtz
Hemholtz
Institut
Fraunhofer-Institut
Fraunhofer-Institut
Max-Planck-Institut
Institute
Max-Plank-Institut