A Visual Introduction to Machine Learning – Part II

r2d3.us

311 points·rrherr··10 comments

Link to part 1: http://www.r2d3.us/visual-intro-to-machine-learning-part-1/

Love to know what kind of tools/techniques/processes were used to generate such beautiful graph/animations....

I've dug through their minified files and from what I can tell other than dependencies they use i.e. Backbone, underscore, jQuery, bootstrap, D3, etc I believe a lot of it is custom coded. It's truly incredible.

Author here – Part 2 uses:

- React for DOM rendering

- D3 for visualization geometry

- RxJS for handling state (i.e. turning scroll position into animated properties)

- PIXI for WebGL bits (i.e. all the little circles moving around)

Ah shoulda mentioned I looked at part 1's minified code and assumed the same for part 2. Thanks for clarifying :)

this is amazing! the animations are a bit lagging sometimes, but the fluidity helps people understand from how single points, once aggregated, can be used to make ML-based decisions.

I noticed on part 1 the graph of SF vs NYC housing elevation is shown in feet, but the text to the side describes the situation in meters. It would probably be better to use the same units between them.

The same issue is found in the 2nd graph. There might be more as it continues.

wow, I really like this format with the ever adapting figures on the right hand side as a way of explaining a concept. Great design.

Yes! We can grasp the concepts much easier like this.

Loved part 1. I have been waiting for an update from r2d3 for a while! Can't wait to check this one out.

Reply on news.ycombinator.com