Algorithm | Dataset | Accuracy (%) | Sensitivity (%) | Specificity (%) |
---|---|---|---|---|
AdaBoost.M1 | Training | 95.33 | 96.64 | 93.06 |
Validation | 87.64 | 88.78 | 86.31 | |
Test | 87.23 | 92.47 | 80.00 | |
Bagging | Training | 84.27 | 88.95 | 77.68 |
Validation | 81.04 | 87.76 | 73.21 | |
Test | 81.93 | 87.63 | 74.07 | |
fastAdaboost | Training | 99.86 | 100.0 | 99.67 |
Validation | 89.84 | 88.78 | 91.07 | |
Test | 84.11 | 89.25 | 77.03 | |
Neyman–Pearson | Training | 97.18 | 99.29 | 94.21 |
Validation | 89.56 | 88.78 | 90.48 | |
Test | 84.73 | 81.72 | 88.89 | |
Stacking | Training | 99.86 | 100.0 | 99.67 |
Validation | 91.76 | 90.31 | 93.45 | |
Test | 86.29 | 88.71 | 82.96 | |
Ensemble strategy | Test | 87.54 | 89.25 | 85.19 |