DiviK: divisive intelligent K-means for hands-free unsupervised clustering in big biological data

BMC Bioinformatics

Table 4 High-level summary of clustering algorithm characteristics

	Spectral clustering	K-means	Spatial clustering	DiviK
Rand index ranks	62.5	71	40.5	36
Dice index ranks	59	75	39	37
EXIMS score ranks	62.5	*38.5*	52.5	56.5
Scalability ranks	72.5	*32.5*	72.5	*32.5*
Total rank	256.5	217	204.5	*162*

Contains sum of ranks for each criterion (lower is better). Top rank of a quality index across clustering methods is presented in bold italics

ISSN: 1471-2105