Skip to content

Latest commit

 

History

History

q_learning_maze

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
images
images
 
 
README.md
README.md
 
 
RL_agent.py
RL_agent.py
 
 
app.py
app.py
 
 
maze_env.py
maze_env.py
 
 
rewards_movements_q_learn.png
rewards_movements_q_learn.png
 
 

README.md

Maze problem with Reinforcement Learning

Maze Environment

The environment can be represented as:

  • States: tiles

  • Actions: Left, Right, Up, Down

  • Reward: +1 for gold state, -1 for black state, 0 for others.

Results

After 50 episodes, the number of movements get converged to the optimal. The reward also goes to 1.

Movement and reward trend