Reinforcement learning (RL) is a term used in the fields of artificial intelligence, automation and Industry 4.0. In short, reinforcement learning describes a method by which computer models learn through trial and error to perform certain tasks better and better - almost like a human being who becomes smarter from experience.
In reinforcement learning (RL), an artificial intelligence is given a task, for example a robot is asked to sort parcels in a factory. At first, the robot does not yet know how to do this correctly. Every time it does something right (for example, puts a parcel in the right place), it receives a "reward", usually in the form of points. If it does something wrong, it gets minus points. Little by little, the robot realises what works and what doesn't.
This principle helps, for example, to make robots work very efficiently in production because they can improve their strategies independently. Reinforcement learning (RL) is also used for digital assistants that make recommendations, for example, in order to better adapt to the needs of users.
Deutsch 
English (UK)