On the Discrepancy between Kleinberg’s Clustering Axioms and k-Means Clustering Algorithm Behavior

This paper performs an investigation of Kleinberg’s axioms (from both an intuitive and formal standpoint) as they relate to the well-known k-mean clustering method. The axioms, as well as a novel variations thereof, are analyzed in Euclidean space. A few natural properties are proposed, resulting in k-means satisfying the intuition behind Kleinberg’s axioms (or, rather, a small, and natural variation on that intuition). In particular, two variations of Kleinberg’s consistency property are proposed, called centric consistency and motion consistency. It is shown that these variations of consistency are satisfied by k-means.