#658 · Primary category: Education & Research

FlappyLearning

flappybird machine-learning neuroevolution

Program learning to play Flappy Bird by machine learning (Neuroevolution)

Project last updated:11/23/23

GitHub Stars

4.0K

Forks

500

Contributors

6

License

MIT

Why we included this project

If you want to see how an evolutionary algorithm can train a neural network without backpropagation, this small project makes it easy to follow. It runs a neuroevolution library in the browser and learns to play Flappy Bird, with code short enough to read through in an afternoon. The standalone Neuroevolution.js file is the useful piece: it exposes options for population size, mutation, elitism, and breeding, so you can watch how each setting affects what the network learns. That makes it a good starting point for classroom demos or for getting comfortable with genetic training before moving to heavier frameworks, even if you would not build production software on it.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category