A Greedy Clustering Algorithm Based on Interval Pattern Concepts and the Problem of Optimal Box Positioning

Table 4

Dependency of Rand index values and the running time for our and -means clustering methods on number of iterations performed (100 program runs for each value). Values of Rand index are presented in terms of medians and interquartile ranges (IQR).

Number of iterations	Clustering method	Rand index median; (adjusted/raw)	Rand index IQR; (adjusted/raw)	Average running time (in seconds)
20	Our method; (with outliers)	0.43/0.73	0.12/0.05	0.8
	Our method; (no outliers)	0.39/0.73	0.10/0.04	0.8
	-means	0.32/0.70	0.21/0.09	0.02
50	Our method; (with outliers)	0.43/0.73	0.08/0.05	2.4
	Our method; (no outliers)	0.39/0.73	0.06/0.03	2.5
	-means	0.27/0.68	0.20/ 0.09	0.05
100	Our method; (with outliers)	0.42/0.74	0.08/0.04	4.2
	Our method; (no outliers)	0.39/0.73	0.06/0.03	4.3
	-means	0.31/0.70	0.20/0.09	0.09