Reinforcement learning

Deep reinforcement learning (deep RL) is the integration of deep learning methods, classically used in supervised or unsupervised learning contexts, with reinforcement learning (RL), a well-studied adaptive control method used in problems with delayed and partial feedback.

Course on deep RL:

https://www.tu-chemnitz.de/informatik/KI/edu/deeprl/

Selected Publications

Winfried Loetzsch, Julien Vitay, and Fred H. Hamker (2017).
Training a deep policy gradient-based neural network with asynchronous learners on a simulated robotic problem.
In: Eibl, M. & Gaedke, M. (Eds.), INFORMATIK 2017. Gesellschaft fuer Informatik, Bonn. (S. 2143-2154)
doi:10.18420/in2017_214