Table 4 Confusion Matrices for the deep learning models
| Â | Inception V3 (Truth) | VGG19 (Truth) | ResNet-50 (Truth) | |||
|---|---|---|---|---|---|---|
Prediction | VRR < 50% | VRR ≥ 50% | VRR < 50% | VRR ≥ 50% | VRR < 50% | VRR ≥ 50% |
Train cohort | ||||||
VRR < 50% | 36 | 22 | 40 | 15 | 42 | 12 |
VRR ≥ 50% | 43 | 83 | 39 | 90 | 37 | 93 |
Validation cohort | ||||||
VRR < 50% | 12 | 14 | 23 | 10 | 14 | 9 |
VRR ≥ 50% | 23 | 31 | 12 | 35 | 21 | 36 |
| Â | EfficientNetB1 (Truth) | EfficientNetB0 (Truth) | EfficientNetb1-fine tune (Truth) | |||
|---|---|---|---|---|---|---|
Prediction | VRR < 50% | VRR ≥ 50% | VRR < 50% | VRR ≥ 50% | VRR < 50% | VRR ≥ 50% |
Train cohort | ||||||
VRR < 50% | 49 | 10 | 53 | 15 | 59 | 7 |
VRR ≥ 50% | 30 | 95 | 26 | 90 | 20 | 98 |
Validation cohort | ||||||
VRR < 50% | 24 | 6 | 21 | 7 | 26 | 8 |
VRR ≥ 50% | 11 | 39 | 14 | 38 | 9 | 37 |