Introduction to Reinforcement Learning

Thomas Gallien
Embedded AI
Silicon Austria Labs
Dienstag, 02.06.2020

In his Science Talk, Thomas Gallien is going to introduce Reinforcement Learning.

Reinforcement learning defines a class of machine learning techniques being able to learn from experience by interacting with an environment. In doing so, we assign a task to an entity called the agent and reinforce it to learn from experience by providing a scalar reward information. This is essentially motivated by the way humans and animals learn. Reinforcement learning algorithms are indeed inspired by the dopamine reward processes of the human brain and, consequently assign machines the role of true learning entities. Within the scope of this talk an introduction in state of the art reinforcement learning will be provided, demonstrating some of the most impressing applications. We will further raise the question how reinforcement learning could find its way in industrial applications.