Skip to main content


Fig. 1 | Journal of Big Data

Fig. 1

From: Efficiency of random swap clustering

Fig. 1

Illustration of the clustering problem using a pigeon-hole principle with dataset S2. The circles represent the location of the ground truth clusters (pigeon holes), and the black dots the prototypes. Centroid index is calculated as how many cluster-level errors (either empty slots or over-crowded slots) there are

Back to article page