S6CI03

Snail jumper game with Neural Networks and Generic Algorithm.

In this project we are going to play the Jumping Snail game with our Neural Network that does not use backpropagation; insted it uses generic algorithm to find the best playing algorithm.

When you want to use the NN mode of the game, it creates 300 agents and starts our Neural Network training. After many generations, it finaly gets to a logical algorithm, but it is not the best algorithm.

How to use?

Clone the project:

git clone https://github.com/amirhnajafiz/neuroevolution-game.git

Install requirements:

pip install -r requirements.txt

Now you can run the game:

python game.py

If you want you can check the generations by the following command:

python history.py