Data Science and Beer

The Bavarian Industry Association for Applied Artificial Intelligence will meet on Monday, October 18, 2021 in Regensburg for two interesting lectures.


  • Causal Inference in Manufacturing (Daniel Grünbaum): “What causes defects in our finished goods? Should we buy better production equipment or more expensive material?” The quickly growing field of causal inference promises answers to urgent causal questions that classical correlation-based statistics cannot solve. Motivated by the quest for root cause analysis in manufacturing, we will get to know the graph-based approach to causal inference, compare it to traditional statistics and briefly look into the subdisciplines of causal discovery and causal estimation. Strategical and computational tools for applying the framework in practice will be presented. The concepts translate seamlessly to a whole range of problems outside of manufacturing that require thorough causal analysis to overcome spurious correlations.
  • Parameterized reinforcement learning in industrial applications – A case study (Heribert Wankerl): Many aspects of human behavior policies in particular situations are acquired via reinforcement learning. The learning theory behind it allows to implement algorithmic agents that manipulate their environments in order to maximize a notion of associated reward. The presentation reveals how to extend common Q-learning to address parameterized Markov decision processes. In semiconductor industry, such methods can help to steer production scheduling, to answer resource allocation questions or assist engineers during product development. Eventually, as an example, we demonstrate how to solve inverse design problems in physics and engineering.

Monday, 10-18-2021, 06.00 p.m. – 09.00 p.m., Restaurant Quetschn (Kallmünzergasse 4, Regensburg)

Further details & registration at:

This is an event of the Industrieverein Bayern für angewandte Künstliche Intelligenz e.V. (Bavarian Industry Association for Applied Artificial Intelligence).