Is it possible to obtain the set of inequalities/rules that would make an observation fall into a specific cluster in k means? By inequalities/rules I mean a set of inequalities for each coordinate to satisfy in order to fall in to an specific cluster.
I know KMeans works by minimizing distances between the centroids and the oberservations, so that is why I am not sure if what I am asking has been implemented.
I'm using the Kmeans module in sklearn.cluster
Something similar was asked here: KMeans: Extracting the parameters/rules that fill up the clusters