For this project, I train an agent to navigate and collect bananas in a square world.
A reward of +1 is provided for collecting a yellow banana, and a reward of -1 is provided for collecting a blue banana. Thus, the goal of the agent was to collect as many yellow bananas as possible while avoiding blue bananas.
The state space has 37 dimensions and contains the agent's velocity, along with ray-based perception of objects around agent's forward direction. Given this information, the agent has to learn how to best select actions. Four discrete actions are available, corresponding to:
0
- move forward.1
- move backward.2
- turn left.3
- turn right.
The task is episodic, and in order to solve the environment, the agent must get an average score of +13 over 100 consecutive episodes.
-
Clone this repo.
-
Setup the python enviroment following next link: udacity/deep-reinforcement-learning
-
Copy the content of the
p1_navigation/
folder from this repo to thep1_navigation/
folder of the udacity/deep-reinforcement-learning repo and replaces or remove existing files. -
Unzip the Banana_Linux.zip file that is located under the
p1_navigation/
folder under the same directory
Opne a jupyter notebook and open the Navigation.ipynb to train or test the agent.
-
For training from zero run all the cells inside the navigation notebook
-
For testing skip the training section inside the notebook
GNU General Public License v3.0