Heuristic Clustering Based on Centroid Learning and Cognitive Feature Capturing

Figure 2

Clustering results of the four synthetic data sets by the new clustering. (a) Result with “Airplane” data set. (b) Result with “Anchor” data set. (c) Result with “Ring” data set. (d) Result with “Swissroll” data set. The four data sets have typical cognitive features, such as density difference, connectedness, and manifolds.

(a)

(b)

(c)

(d)