Add new comment

First, I found this post extremely insightful and interesting. I used data project in a project not that long ago and it was awesome !

Here is it running with a totally different, dimensionality reduced set of data. In this case, it was PCA, then SVD I think.

The data in that case is some strange clustering (or lack there of) of ted-talks in 2013...

I took your idea and tried to do something similar here:


Submitted by xander johnson on

RSS Feed