Table 3.
Results of classification incorporating Xception deep features applied on the HAM10000 dataset.
| Classifiers | Accuracy (%) |
Recall (%) |
Precision (%) |
F1 Score | FPR | Time (sec) |
|---|---|---|---|---|---|---|
| Narrow NN | 98.9 | 98.97 | 98.88 | 98.17 | 1.03 | 469.27 |
| Medium NN | 98.9 | 99.01 | 99.02 | 99.01 | 0.99 | 261.74 |
| Wide NN | 98.9 | 99.02 | 99.05 | 99.03 | 0.98 | 334.87 |
| Bilayered NN | 98.8 | 98.92 | 98.92 | 98.92 | 1.08 | 500.02 |
| Quadratic SVM | 99.0 | 99.1 | 99.17 | 99.12 | 0.88 | 1009.5 |
| Cubic SVM | 99.1 | 99.18 | 99.22 | 99.18 | 0.82 | 1017.9 |
| Coarse Gaussian SVM | 98.7 | 98.77 | 98.44 | 99 | 1.23 | 1931 |
| Fine KNN | 99.1 | 99.15 | 99.2 | 99.16 | 0.85 | 2002 |
| Ensemble Bagged Tree | 98.8 | 98.9 | 98.9 | 98.9 | 1.1 | 327.58 |
| Fine Tree | 94.9 | 95.15 | 95.15 | 95.18 | 4.85 | 100.25 |
Bold show the significant value.