Ezgii/Escaping-maze-using-RL - Githubissues

Ezgii / Escaping-maze-using-RL

Escaping from a maze using Reinforcement Learning techniques.

0 stars 0 forks source link

maze policyiteration qlearning reinforcement-learning sarsa

readme

In this project, the optimal policy to escape a maze is found using model-based (Policy Iteration) and model-free (Q-learning and SARSA) reinforcement learning techniques.

Start state : top-left corner, Goal state : bottom-right corner

Policy Iteration result:

Q-Learning result:

Q-learning

SARSA result:

SARSA