Table 6.
Computational (CPU) time (s) for 3 algorithms to calculate mixed-distance dissimilarity.
| Dissimilarity Algorithm |
Dissimilarity Time (s) | Clustering Time (s) |
|
|---|---|---|---|
| HC1 | PAM2 | ||
| DAISY | 372.461 ± 983.599 | 0.105 ± 0.142 | 1.623 ± 3.611 |
| Mercator | 99.859 ± 139.127 | 0.097 ± 0.13 | 1.598 ± 3.46 |
| Supersom3 | 0.533 ± 0.794 | - | - |
1
Agglomerative hierarchical clustering with Ward’s criterion.
2
Partitioning Around Medoids.
3
Kohonen self-organizing maps and their related mixed-distance implementation calculate dissimilarity and cluster in a single process.