Table 2.
Accuracy, F1-score and Std. dev. results of validation and test for the COVID-19 dataset
| Validation | Test | |||||||
|---|---|---|---|---|---|---|---|---|
| Acc. (mean) | Acc. (max) |
F1-score | Std. dev. | Acc. (mean) | Acc. (max) |
F1-score | Std. dev. | |
| AlexNet | 90.065 | 93.638 | 90.254 | 2.627 | 87.089 | 90.601 | 87.305 | 2.590 |
| DarkNet-19 | 93.189 | 95.269 | 93.292 | 1.833 | 91.149 | 94.125 | 91.309 | 1.526 |
| Inception-v3 | 89.233 | 93.964 | 89.291 | 3.610 | 86.436 | 90.601 | 86.631 | 3.171 |
| MobileNet | 82.007 | 86.134 | 81.716 | 3.052 | 81.312 | 85.248 | 81.333 | 2.470 |
| ResNet-18 | 91.313 | 93.638 | 91.418 | 1.460 | 88.962 | 91.253 | 89.115 | 1.793 |
| ShuffleNet | 88.679 | 94.454 | 88.665 | 4.035 | 86.449 | 90.339 | 86.535 | 3.075 |
| COVID-CCD-Net (AlexNet) | 96.354 | 97.879 | 96.460 | 1.166 | 96.044 | 98.172 | 96.138 | 0.927 |
| COVID-CCD-Net (DarkNet-19) | 97.553 | 98.532 | 97.654 | 0.631 | 97.369 | 98.303 | 97.458 | 0.496 |
| COVID-CCD-Net (Inception-v3) | 95.718 | 97.553 | 95.830 | 1.139 | 94.791 | 96.736 | 94.900 | 1.736 |
| COVID-CCD-Net (MobileNet) | 95.057 | 97.227 | 95.179 | 1.502 | 94.608 | 98.042 | 94.718 | 1.816 |
| COVID-CCD-Net (ResNet-18) | 97.977 | 98.532 | 98.063 | 0.301 | 98.107 | 98.695 | 98.158 | 0.353 |
| COVID-CCD-Net (ShuffleNet) | 96.223 | 96.900 | 96.312 | 0.424 | 96.730 | 98.172 | 96.805 | 0.623 |
| Quasi-Newton-based AlexNet | 94.043 | 96.574 | 94.205 | 4.123 | 91.879 | 93.994 | 92.099 | 3.038 |
| Quasi-Newton-based DarkNet-19 | 96.807 | 97.226 | 96.914 | 0.373 | 94.072 | 95.430 | 94.225 | 1.399 |
| Quasi-Newton-based Inception-v3 | 94.266 | 95.432 | 94.331 | 0.868 | 91.051 | 93.733 | 91.179 | 2.554 |
| Quasi-Newton-based MobileNet | 93.541 | 95.595 | 93.653 | 1.723 | 89.713 | 92.298 | 89.925 | 1.990 |
| Quasi-Newton-based ResNet-18 | 95.834 | 96.248 | 95.916 | 0.511 | 93.019 | 94.125 | 93.124 | 0.869 |
| Quasi-Newton-based ShuffleNet | 95.050 | 95.759 | 95.142 | 1.055 | 90.551 | 93.081 | 90.731 | 2.336 |