Deutsch Intern
    Center for Artificial Intelligence and Data Science

    Chair for Reinforcement Learning and Computational Decision-Making

    Another central area of machine learning is addressed by the Chair of Reinforcement Learning and Computational Decision-Making. The methods subsumed under this term are based on the natural learning behaviour of humans.  An agent independently learns a strategy based on feedback from its environment with the goal of maximising rewards received over time. Reinforcement learning allows strategic action, i.e. an independent reaction to new circumstances, the anticipation of possible future actions and their consideration in current decisions, and thus forms an important basis for self-learning, autonomous artificial intelligence algorithms.
    In economics, these methods play an important role in the subfield of operations research.  Thus, intelligent planning in complex techno-economic systems is to be a central field of application of the chair's work. The main focus here is on dealing with systems that are only partially observable. The strategies and procedures determined by reinforcement learning are typically not directly interpretable and therefore do not provide any conclusions about the underlying causal relationships. Inverse reinforcement learning methods address this gap and thus offer an opportunity to methodically link machine learning, intelligent planning and econometrics.
    The Chair will benefit from the theoretical developments, e.g. of the Chair of Theory of Machine Learning, just as other fields of application will benefit from the new optimisation methods of its research work.

    ongoing appointment procedure

    • estimated date of appointment: 2022
    • application deadline: 10 May 2021
    • job advertisment: go.uniwue.de/w3-rlcdm-en
    • level: W3 (full professorship)