|
No. | Name | Formula | Optimal clustering number | References |
|
1 | CH index | | Maximum value of the index | Caliński and Harabasz [22] |
2 | C Index | | Minimum value of the index | Hubert and Levin [23] |
3 | Gamma index | | Maximum value of the index | Baker and Hubert [24] |
4 | DB index | | Minimum value of the index | Davies and Bouldin [25] |
5 | Hartigan index | | Maximum difference between hierarchy levels of the index | Hartigan [26] |
6 | Tau index | | Maximum value of the index | |
7 | Scott index | | Maximum difference between hierarchy levels of the index | Scott and Symons [27] |
8 | Marriot index | | Maximum value of second differences between levels of the index | Marriot [28] |
9 | Friedman index | | Maximum difference between hierarchy levels of the index | Friedman and Rubin [29] |
10 | Rubin index | | Minimum value of second differences between levels of the index | Friedman and Rubin [29] |
11 | KL index |
| Maximum value of the index | Krzanowski and Lai [30] |
12 | Silhouette index |
,
| Maximum value of the index | Rousseeuw [31] |
13 | Dunn index |
| Maximum value of the index | Dunn [32] |
14 | SD index |
σ = (VAR (V1), VAR (V2), …, VAR(VP))
Dmax = max(||ck − cz||) ∀k, z ∈ {1, 2, 3, …, q} Dmin = min(||ck − cz||) ∀k, z ∈ {1, 2, 3, …, q}
| Minimum value of the index | Halkidi et al. [33] |
|