Skip to main content
. 2021 May 4;180:115141. doi: 10.1016/j.eswa.2021.115141

Table 2.

The calculated performance metrics of the proposed model.

DenseNet_201
TP FP FN TN Accuracy(%) Precision (%) Recall(%) F1-score
Fold-1 61 7 5 146 94.52 89.71 92.42 91.04
Fold-2 58 3 8 150 94.98 95.08 87.88 91.34
Fold-3 60 2 6 151 96.35 96.77 90.91 93.75
Fold-4 61 5 5 148 95.43 92.42 92.42 92.42
Fold-5 62 3 4 150 96.80 95.38 93.94 94.66
Total 302 20 28 745 95.62 93.79 91.52 92.64
ResNet_18
TP FP FN TN Accuracy(%) Precision(%) Recall(%) F1-score
Fold-1 58 9 8 144 92.24 86.57 87.88 87.22
Fold-2 59 13 7 140 90.87 81.94 89.39 85.51
Fold-3 61 8 5 145 94.06 88.41 92.42 90.37
Fold-4 61 11 5 142 92.69 84.72 92.42 88.41
Fold-5 60 3 6 150 95.89 95.24 90.91 93.02
Total 299 44 31 721 93.15 87.17 90.61 88.86
SqueezeNet
TP FP FN TN Accuracy(%) Precision(%) Recall(%) F1-score
Fold-1 58 11 8 142 91.32 84.06 87.88 85.93
Fold-2 60 9 6 144 93.15 86.96 90.91 88.89
Fold-3 59 10 7 143 92.24 85.51 89.39 87.41
Fold-4 61 12 5 141 92.24 83.56 92.42 87.77
Fold-5 62 5 4 148 95.89 92.54 93.94 93.23
Total 300 47 30 718 92.97 86.46 90.91 88.63

TP: True Positive; FP:False Positive; FN: False Negative; TN: False Positive.