Table 8 Comparison of specificity, AUC-ROC, and sensitivity.
Model | Specificity (%) | AUC-ROC (%) | Sensitivity (%) |
|---|---|---|---|
CNN | 88.52 | 88.23 | 87.77 |
ResNet | 90.87 | 90.87 | 89.92 |
DenseNet | 91.38 | 91.56 | 90.56 |
VGG | 90.79 | 90.21 | 89.84 |
AutoEncoder-LSTM | 91.11 | 91.11 | 90.98 |
Hybrid ResNet-LightGBM | 92.48 | 92.18 | 91.87 |
Inception-XGBoost | 92.05 | 91.93 | 91.41 |
EfficientNet-gradient boosting | 92.64 | 92.66 | 91.73 |
Capsule network | 91.86 | 92.05 | 91.32 |
Proposed EDHL (Inception-GB) | 95.98 | 98.75 | 97.98 |
