K-Means clustering and scale variables
What mean is the 'scale variables' option in K-Means clustering working with? Is it the mean for each dimension or ...? For example, the image below has the cluster means for one dimension. The values on the left are raw and the values on the right are scaled. Why does 0.801 and 0.287 scale to be positive when 0.728 standardizes to be negative?
Comments
I've asked, hope to learn from the team members soon
Cheers,
E.J.