Decision tree-based method for integrating gene expression, demographic, and clinical data to determine disease endotypes

Table 6 Top ranking 67 covariate clustering methods from clusterSim

Index metric	Index value	Distance measure	Clustering method	No. of clusters
Silhouette	0.6692	Generalized Distance Measure	Partitioning Around Medoids	2
Baker & Hubert	0.9122	Chebyschev	Hierarchical - Single linkage	2
Hubert & Levine	0.0279	Generalized Distance Measure	Partitioning Around Medoids	24
		Generalized Distance Measure	Hierarchical - Average linkage	8
		Generalized Distance Measure	Hierarchical - Average linkage	14
		Generalized Distance Measure	Hierarchical - Average linkage	13

The optimal distance measure and clustering method using three separate indices are shown along with the associated index value in each case. Where no index metric or value is given, an attempt was made to create more informative clusters rather than optimize a clustering index.

ISSN: 1752-0509