Dissimilarity
Fundamental to clustering is a measure of similarity (or dissimilarity) of the objects
being clustered. Clustering seeks to group together those objects which are most
similar (or least dissimilar). The within-cluster dissimilarity can form the basis for
a loss function that clustering seeks to minimize, in much the same way that linear
regression seeks to minimize the sum of squared distances between observed and
fitted values.
Common measures of dissimilarity for continuous data include Euclidean distance