Research Article
Clustering Ensemble Model Based on Self-Organizing Map Network
Table 5
The average performance of ten runs for each dataset measured by NMI.
| ā | CO-Average | ONCE-Average | DSCE | ACE | DICLENS | MCLA | Cascaded SOM |
| Iris | 0.751 | 0.752 | 0.763 | 0.766 | 0.757 | 0.749 | 0.757 | Wine | 0.428 | 0.428 | 0.432 | 0.429 | 0.427 | 0.429 | 0.434 | Thyroid | 0.434 | 0.473 | 0.480 | 0.531 | 0.501 | 0.418 | 0.492 | Mfeatures | 0.479 | 0.479 | 0.479 | 0.478 | 0.468 | 0.475 | 0.480 | Glass | 0.712 | 0.725 | 0.725 | 0.726 | 0.617 | 0.728 | 0.731 | BCW | 0.750 | 0.749 | 0.750 | 0.751 | 0.742 | 0.751 | 0.748 | Soybean | 0.717 | 0.723 | 0.756 | 0.712 | 0.822 | 0.717 | 0.785 | Ionosphere | 0.122 | 0.124 | 0.128 | 0.123 | 0.119 | 0.124 | 0.126 |
|
|
The bold values mean the best NMI value of the different algorithms for each dataset.
|