A collaborative list of interactive Machine Learning, Deep Learning and Statistics websites. (Started by Piotr Migdał, but anyone is encouraged to contribute!)

Check websites.yaml. I am developing a simple no-build Vue.js website. Right now is in primordial stage (Web 1.0!) and contributions will be appreciated. Here it is.

Feel invited to Pull Request other interactive visualizations! :)

Still I am thinking what is the best criterion.

For sure things that are front-end (i.e. JavaScript within browser). For things using backend (when you can see solution, but it uses some PyTorch/TF/etc code on a server) I am still debating, but I lean on being more inclusive. In this context:

• make sure it has some didactic value (otherwise ALL services using ML would qualify)
• add backend-dependent in uses

Strong preference for open-source solutions (so people can reuse it and learn from code), though it is not a requirement. Though, mention repo and open source license only when it is directly relevant (vs additional materials such as exercises for a book, or Python algorithm).

Q: I saw its current website looks bare. Can I help?

A: Yes! (Myself I am not a frontend developer. I play with it for fun, but collaboration is welcomed!)

• Explorable Explanations
• Distill
• Explained Visually

Inspirations for collecting and displaying content:

• Science-based games - a collaborative list - a list I started (maybe I will turn it int something interactive as well)
• Kaggle Past Solutions - a searchable compilation of Kaggle past solutions
  • source: EliotAndres/kaggle-past-solutions
• D3 Discovery - finding D3 plugins with ease
  • source: https://github.com/wbkd/d3-discovery