Table 3 Validation Results: The mean classification error, normalized mutual information (NMI) and stability, on all datasets, are shown, measuring the agreement between the clusters resulting from an approach and the real patient classification
From: MVDA: a multi-view genomic data integration methodology
| Feature | Integration | Algorithm | Error | NMI | Stability | |
|---|---|---|---|---|---|---|
| Single View | All Feature | - | Ward | 30,08 % | 26 % | 86 % |
| - | Kmeans | 30,93 % | 25 % | 51 % | ||
| - | Pamk | 30,75 % | 24 % | 94 % | ||
| Selected Prototype | - | Ward | 30,72 % | 26 % | 89 % | |
| - | Kmeans | 30,36 % | 25 % | 52 % | ||
| - | Pamk | 30,78 % | 24 % | 96 % | ||
| Multi-View | All Feature | Early Integration | Tw-kmeans | 37,10 % | 24 % | 69 % |
| All Feature | Intermediate Integration | SNF | 30,83 % | 22 % | 83 % | |
| All Feature in Cluster of Selected Prototype | Intermediate Integration | SNF | 31,31 % | 18 % | 82 % | |
| Selected Prototype | Late Integration unsupervised | MF/GLI | 27,47 % | 28 % | 85 % | |
| Selected Prototype | Late Integration semi-supervised | MF/GLI | 6,30 % | 63 % | 84 % |