Clustering Analysis of Multivariate Data: A Weighted Spatial Ranks-Based Approach

Table 2

Comparison of different clustering approaches applied to the financial dataset.

Clustering method	No. of clusters	Cluster sizes	H	Purity	Entropy	ARI
WSR	2	53, 50	0.03	0.97	0.16	0.89
GMM (mclust) “BIC”	3	50, 15, 38	0.15	0.94	0.14	0.73
K-means “CH index”	2	46, 57	0.03	0.97	0.19	0.89
HDDC “BIC”	3	8, 45, 50	0.09	0.98	0.09	0.82
MixtPPCA “BIC”	2	53, 50	0.03	0.97	0.19	0.89
PAM “silhouette width”	2	57, 46	0.03	0.97	0.19	0.89
DBSCAN	1	102 (1 noise point)	0.03	0.54	0.99	−0.003
KMD	2	46, 57	0.03	0.97	0.19	0.89
FCM	2	46, 57	0.03	0.97	0.19	0.89
GG	2	53, 50	0.03	0.97	0.16	0.89
DDC	3	11, 46, 46	0.12	0.97	0.14	0.76
SNN	3	47, 40, 16	0.16	0.93	0.15	0.71
densityClust	1	103	0	0.54	0.99	0

Results are based on the mean of 1000 repetitions.