Reinforcement Learning - Maze

environment: maze, on which a rat is trying to catch cheese
agent: neural network

TODO: more information (q-learning, Bellmann-Equation, etc.)

A few examples

8x8 maze

20x20 maze

9x20 maze