TABLE 7.
Performance metric of the ACRIMA dataset using different CNN models.
| CNN model | AUC | ACC | SEN | SP | PRE | FM | GM |
|---|---|---|---|---|---|---|---|
| ResNet50 | 1.0000 | 0.9986 | 0.9975 | 1.0000 | 1.0000 | 0.9987 | 0.9987 |
| VGG19 | 0.9989 | 0.9858 | 0.9798 | 0.9935 | 0.9949 | 0.9873 | 0.9866 |
| AlexNet | 0.9890 | 0.9461 | 0.9217 | 0.9773 | 0.9812 | 0.9505 | 0.9491 |
| Dns201 | 0.9998 | 0.9901 | 0.9848 | 0.9968 | 0.9974 | 0.9911 | 0.9908 |
| IncRes | 0.9984 | 0.9291 | 0.8737 | 1.0000 | 1.0000 | 0.9326 | 0.9347 |
| Fusion | — | 0.9957 | 0.9962 | 1.0000 | 0.9952 | 0.9962 | 0.9957 |
Bold values are showing the model results after applying classifier fusion operation.