From: Implementation of multiple-instance learning in drug activity prediction
Data set | Dissimilarity measure | Training set | Test set | ||
---|---|---|---|---|---|
 |  | Accuracy | MCC | Accuracy | MCC |
I | Soergel | 0.972 | 0.944 | 0.854 | 0.714 |
 | Dice | 0.979 | 0.959 | 0.825 | 0.653 |
 | Manhattana | 0.941 | 0.881 | 0.861 | 0.725 |
 | Rogers-Tanimoto | 0.961 | 0.923 | 0.861 | 0.725 |
II | Soergel | 0.965 | 0.933 | 0.860 | 0.725 |
 | Dice | 0.965 | 0.933 | 0.868 | 0.745 |
 | Manhattana | 0.978 | 0.956 | 0.904 | 0.807 |
 | Rogers-Tanimoto | 0.973 | 0.946 | 0.897 | 0.793 |
III | Soergel | 0.989 | 0.977 | 0.846 | 0.706 |
 | Dice | 0.989 | 0.977 | 0.855 | 0.717 |
 | Manhattan | 0.979 | 0.954 | 0.838 | 0.686 |
 | Rogers-Tanimotoa | 0.947 | 0.885 | 0.846 | 0.711 |
IV | Soergel | 0.904 | 0.823 | 0.667 | 0.301 |
 | Dice | 0.904 | 0.823 | 0.635 | 0.307 |
 | Manhattan | 0.957 | 0.918 | 0.714 | 0.433 |
 | Rogers-Tanimotoa | 0.898 | 0.811 | 0.794 | 0.584 |