TABLE 2.
Comparison of accuracy and AUC of each model for multiclass classification.
| Model | Dataset | Accuracy | AUC |
|---|---|---|---|
| VGG 19 | Day 1 | 0.9825 | 1.0000 |
| Day 2 | 1.0000 | 1.0000 | |
| Day 3 | 0.8814 | 0.9957 | |
| Day 5 | 0.8833 | 0.9957 | |
| Day 7 | 0.9661 | 0.9849 | |
| Day 10 | 0.9649 | 1.0000 | |
| Day 13 | 0.9184 | 1.0000 | |
| Day 15 | 0.9821 | 1.0000 | |
| Overall | 0.9474 | 0.9928 | |
| Inception V3 | Day 1 | 0.9474 | 0.9935 |
| Day 2 | 0.9831 | 0.9978 | |
| Day 3 | 0.8305 | 0.9769 | |
| Day 5 | 0.8333 | 0.9667 | |
| Day 7 | 0.9322 | 0.9654 | |
| Day 10 | 0.9649 | 1.0000 | |
| Day 13 | 1.0000 | 1.0000 | |
| Day 15 | 1.0000 | 1.0000 | |
| Overall | 0.9342 | 0.9899 | |
| ResNet18 | Day 1 | 0.9649 | 0.9995 |
| Day 2 | 0.9661 | 0.9983 | |
| Day 3 | 0.8305 | 0.9871 | |
| Day 5 | 0.9500 | 0.9987 | |
| Day 7 | 0.9492 | 0.9771 | |
| Day 10 | 0.9649 | 1.0000 | |
| Day 13 | 1.0000 | 1.0000 | |
| Day 15 | 0.9107 | 0.9776 | |
| Overall | 0.9408 | 0.9925 | |
| ResNet 50 | Day 1 | 0.9474 | 0.9991 |
| Day 2 | 0.9661 | 0.9987 | |
| Day 3 | 0.8644 | 0.9833 | |
| Day 5 | 0.9167 | 0.9867 | |
| Day 7 | 0.9661 | 0.9893 | |
| Day 10 | 0.9474 | 1.0000 | |
| Day 13 | 1.0000 | 1.0000 | |
| Day 15 | 0.9821 | 0.9990 | |
| Overall | 0.9474 | 0.9936 |