Table 2 Validity indices used to estimate the optimum number of data clusters

Hierarchical clusters

ICT

K-means clusters

ADSM

9

2.35

3

0.755

16

2.3

4

0.678

35

2.25

5

0.605

38

2.2

6

0.509

41

2.15

7

0.497

44

2.1

8

0.490

48

2.05

9

0.472

50

2

10

0.485

CGI attribute data

Hierarchical clusters

ICT

K-means clusters

ADSM

1

2.7

3

0.221

8

2.65

4

0.233

9

2.6

5

0.184

9

2.55

6

0.177

9

2.5

7

0.169

16

2.45

8

0.145

24

2.4

9

0.134

31

2.35

10

-0.348