A foray into genetic algorithms. I wrote a quick pendulum balancing game, and then wrote a neural network that tried to get highscores in it. This was trained with a genetic algorithm. Running this is more of a demo of this training technique than a game.
- Install all the requirements from the requirements.txt
- To play the game yourself, you can run
python game.py
- To see the genetic algorithm learn, you can run
python trainer.py
. If you want to edit the variables such asgen_size
, go into thetrainer.py
and fiddle all you want
- Clicking on the window will spawn your mouse cursor as a box with collision if you want to mess around with the networks.
- Report any other errors under issues, and I may or may not get onto to solving it.
- Enjoy!!!