Reinforcement Learning (RL) deals with optimization of sequential decision making processes. It takes its sources into animal psychology, dynamic programming and temporal difference learning. It relies on (Partially Observable) Markov Decision Processes ((PO)MDP) and combines bio-inspired and statistical machine learning. The IMS group is both interested in the mathematical foundations of RL and its biological explanation.
Besides, IMS also applies Reinforcement learning to spoken dialogue manangement optimization.