Table 2 Average Silhouette index scores over 20 runs of different algorithms for the five gene expression data sets
From: Combining Pareto-optimal clusters using supervised learning for identifying co-expressed genes
Algorithm | Sporulation | Cell Cycle | Thaliana | Serum | Rat CNS | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Â | K | s(C) | K | s(C) | K | s(C) | K | s(C) | K | s(C) |
MOGA-SVM (RBF) | 6 | 0.6283 | 5 | 0.4426 | 4 | 0.4312 | 6 | 0.4154 | 6 | 0.5127 |
MOGA (without SVM) | 6 | 0.5794 | 5 | 0.4392 | 4 | 0.4011 | 6 | 0.3947 | 6 | 0.4872 |
MOGA crisp -SVM (RBF) | 6 | 0.5971 | 5 | 0.4271 | 4 | 0.4187 | 6 | 0.3908 | 6 | 0.4917 |
FCM | 7 | 0.4755 | 6 | 0.3872 | 4 | 0.3642 | 8 | 0.2995 | 5 | 0.4050 |
SGA | 6 | 0.5703 | 5 | 0.4221 | 4 | 0.3831 | 6 | 0.3443 | 6 | 0.4486 |
Average linkage | 6 | 0.5007 | 4 | 0.4388 | 5 | 0.3151 | 4 | 0.3562 | 6 | 0.4122 |
SOM | 6 | 0.5845 | 6 | 0.3682 | 5 | 0.2133 | 6 | 0.3235 | 5 | 0.4430 |
CRC | 8 | 0.5622 | 5 | 0.4288 | 4 | 0.4109 | 10 | 0.3174 | 4 | 0.4423 |