Hosted at machinelearning.tadeaspetak.com.
This is a short visualization demo of the most basic mathematical principles behind machine learning. I wrote it to support my 15-minute talk during a competence day at Jayway in March, 2016. While it's not the most organised demo in the world, not to mention it's not even close to being responsive or anything like that, it fulfills its purpose: it visualizes cost function and gradient descent principles.
It's written in isomorphic React. I wanted to give that a try and I am fairly happy with the setup and results. I wrote the visualization itself in pure d3 which was fun since I had never used the library before. It's pretty awesome.