Python Idle
You've been hired by a robotics startup building autonomous warehouse robots. Before deploying in the real warehouse, you need to prove the concept in simulation. A 4x4 grid world represents a simplified floor plan: the robot starts in one corner and must learn to navigate to the goal in the opposite corner while avoiding a dangerous pit. No labeled data here — the agent must learn entirely from trial and error. Welcome to reinforcement learning.
~25 minscenario
Loading Python runtime...
Goals: 4 tests
Q-table should have shape (16, 4)
should train for at least 100 episodes
Q-table should have non-zero values after training
Goal state should have the highest Q-value
Python loading...