Table 3 Summary results for six datasets comparing accuracy and affinity for the single measure and dual measure framework
From: Finding reproducible cluster partitions for the k-means algorithm
| Â | Single Measure | Dual Measure | ||
|---|---|---|---|---|
Dataset | Accuracy (Std. Dev) | Affinity | Accuracy (Std. Dev) | Affinity |
Artificial 500 | 0.7758 (0.038) | 0.922 | 0.7701 (0.015) | 0.931 |
Artificial 1,000 | 0.9263 (0.037) | 0.994 | 0.7773 (0.015) | 0.980 |
Artificial 2,000 | 0.7332 (0.058) | 0.888 | 0.7345 (0.003) | 0.992 |
Artificial 5,000 | 0.9079 (0.089) | 0.937 | 0.961 (0.008) | 0.989 |
Artificial 10,000 | 0.9929 (0.032) | 0.993 | 0.9994 (0.001) | 0.999 |
Cardiotocography | 0.3655 (0.017) | 0.792 | 0.3775 (0.002) | 0.983 |