Table 2 Performance metrics of ResNet50 and ResNet50v2 with optimizers SAM and SGD continuity analysis.
CNN | Optimizer | Accuracy | Precision | Recall | F1 score | AUC |
|---|---|---|---|---|---|---|
SD | SD | SD | SD | SD | ||
95% CI | 95% CI | 95% CI | 95% CI | 95% CI | ||
ResNet50 | SAM | 0.754 | 0.755 | 0.754 | 0.753 | 0.832 |
0.005 | 0.008 | 0.008 | 0.008 | 0.006 | ||
0.753–0.756 | 0.752–0.757 | 0.751–0.757 | 0.750–0.755 | 0.829–0.834 | ||
ResNet50 | SGD | 0.754 | 0.754 | 0.754 | 0.752 | 0.830 |
0.007 | 0.008 | 0.008 | 0.008 | 0.006 | ||
0.752–0.757 | 0.752–0.757 | 0.751–0.757 | 0.750–0.755 | 0.827–0.832 | ||
ResNet50v2 | SAM | 0.766 | 0.766 | 0.765 | 0.775 | 0.843 |
0.007 | 0.006 | 0.006 | 0.013 | 0.005 | ||
0.764–0.769 | 0.764–0.768 | 0.763–0.767 | 0.771–0.780 | 0.842–0.845 | ||
ResNet50v2 | SGD | 0.765 | 0.765 | 0.765 | 0.767 | 0.842 |
0.006 | 0.006 | 0.006 | 0.013 | 0.005 | ||
0.763–0.768 | 0.763–0.767 | 0.762–0.767 | 0.762–0.772 | 0.840–0.844 |
