Dot, dot, comma, dash …
With the help of reinforcement learning, algorithms can learn a certain behaviour through feedback. The actions of a software agent or a robot are assessed based on their usefulness in achieving a specific behavioural goal and the system receives rewarding or punishing feedback. Based on this, the system optimises its behaviour in many iterations. Working with this box, students can train a robot to draw a small house on a display.
Unplugged materials or an online version of this box are not yet available!
1/4
2/4
3/4
4/4