README.md

A_maze

An RL agent that solves mazes using q learning (q tables) and bellman optimality equation.

So get A_MAZE_ed

Here is the maze_layout

Here is the solution after 1000 episodes of training

Here are some interesting plots