I want to do unsupervised learning. As I understand with such learning we don't know the clusters before, right? I read about k-means alghoritm, followed mainly these two articles:
And maybe this sounds stupid but this is some kind of magic for me, that following the example from the second article suddenly out of nothing "Python" knows the clusters.
My problem is to understand how it works from inside. I have around 100 text documents (resumes) and would like to play with this alghoritm - maybe start with 2 clusters - experienced/notExperienced but don't want just to copy the code from tutorials.