Clustering and how to begin
Can one talk about properties that are sufficient (or both necessary and sufficient) for natural algorithms to succeed?
Before diving into details on how to optimize the clustering, let’s go back and think the problem that we need to solve. “Clustering” has to do with the grouping of similar or related objects.
First of all, even before start thinking of clustering the problem is: Do the data have a tendency to form clusters? If the…