Detecting Spatial Clusters via a Mixture of Dirichlet Processes

Table 2

Kmeans simulation assessments for three foci-level clusters.

Scenario:	Median Num. of Clusters  (SD)	Cluster index	Average Sensitivity  (SD)	Average Specificity  (SD)	Average Accuracy (%)  (SD)
Normal	5 (0.95)	1	0.73 (0.29)	0.98 (0.09)	0.7 (0.12)
		2	0.72 (0.3)	0.98 (0.08)
		3	0.67 (0.26)	1 (0)
Chi-squared	10 (0.5)	1	0.58 (0.19)	0.99 (0.01)	0.5 (0.09)
		2	0.86 (0.28)	0.92 (0.03)
		3	0.16 (0.08)	0.98 (0.11)
Large 1	5 (0.98)	1	0.68 (0.33)	1 (0.04)	0.7 (0.13)
		2	0.76 (0.26)	0.96 (0.12)
		3	0.67 (0.26)	1 (0)
Large 2	7 (1.29)	1	0.61 (0.27)	0.99 (0.06)	0.58 (0.11)
		2	0.61 (0.27)	0.99 (0.06)
		3	0.54 (0.27)	1 (0)
Large 3	9 (0.94)	1	0.49 (0.19)	1 (0)	0.46 (0.08)
		2	0.47 (0.17)	1 (0)
		3	0.43 (0.24)	1 (0)
Large 4	10 (0.57)	1	0.43 (0.16)	1 (0.01)	0.41 (0.07)
		2	0.42 (0.14)	1 (0.01)
		3	0.38 (0.2)	1 (0)

SD: standard deviation across 100 MC replicates.