Home

Awesome

REINFORCE-DL4S

Implements the REINFORCE algorithm in DL4S for an agent in a grid world.

Setup

The environment is a 2D grid world with an exit and randomly generated obstacles.

The goal of the agent is to reach the exit from a random starting position.

X: Obstacle
O: Exit
•: Agent

 -------------------- 
|O             X     |
|               X    |
|                    |
|X                   |
|          X         |
|         X     X    |
|            •       |
| X                  |
|             X      |
|  X        X        |
 --------------------