Table 3 Baseline metric evaluation result before augmentation.
From: Minimal sourced and lightweight federated transfer learning models for skin cancer detection
Algorithm | Accuracy | Validation accuracy | Precision | Validation precision | Recall | Validation recall | Loss | Validation loss |
|---|---|---|---|---|---|---|---|---|
DenseNet201 | 74.28 | 64.84 | 82.99 | 70.06 | 62.34 | 58.48 | 1.12 | 1.81 |
EfficientNetB3 | 87.86 | 69.07 | 91.51 | 72.39 | 82.29 | 64.41 | 0.69 | 1.48 |
EfficientNetV2S | 84.78 | 68.23 | 89.37 | 71.43 | 78.47 | 65.68 | 0.82 | 1.54 |
MobileNetV2 | 77.88 | 59.33 | 85.48 | 63.73 | 67.43 | 55.09 | 1.08 | 1.79 |
VGG16 | 35.02 | 22.89 | 64.78 | 45.24 | 3.03 | 16.11 | 1.91 | 2.04 |
VITB16 | 58.78 | 63.14 | 69.7 | 74.16 | 44.41 | 55.94 | 1.19 | 0.99 |
