Table 6.
Models/Folds(F) | Performance Evaluation based on Four Metrics | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
TP | TN | FP | FN |
Test set (%) Acc |
Rec | Spe | Pre | F1 | ||
AlexNet | F-1 | 235 | 643 | 134 | 34 | 83.94 | 87.36 | 82.75 | 63.69 | 73.67 |
F-2 | 238 | 643 | 134 | 31 | 84.23 | 88.48 | 82.75 | 63.98 | 74.26 | |
F-3 | 240 | 660 | 117 | 29 | 86.04 | 89.22 | 84.94 | 67.23 | 76.68 | |
F-4 | 240 | 660 | 117 | 29 | 86.04 | 89.22 | 84.94 | 67.23 | 76.68 | |
F-5 | 240 | 660 | 117 | 29 | 86.04 | 89.22 | 84.94 | 67.23 | 76.68 | |
Average | 85.26 | 88.7 | 84.07 | 65.87 | 75.59 | |||||
GoogleNet | F-1 | 245 | 688 | 89 | 24 | 89.20 | 91.08 | 88.55 | 73.35 | 81.26 |
F-2 | 245 | 698 | 79 | 24 | 90.15 | 91.08 | 89.83 | 75.62 | 82.63 | |
F-3 | 249 | 704 | 73 | 20 | 91.11 | 92.57 | 90.6 | 77.33 | 84.26 | |
F-4 | 249 | 704 | 73 | 20 | 91.11 | 92.57 | 90.6 | 77.33 | 84.26 | |
F-5 | 249 | 704 | 73 | 20 | 91.11 | 92.57 | 90.6 | 77.33 | 84.26 | |
Average | 90.54 | 91.97 | 90.04 | 76.19 | 83.34 | |||||
DenseNet121 | F-1 | 249 | 709 | 68 | 20 | 91.59 | 92.57 | 91.25 | 78.55 | 84.98 |
F-2 | 251 | 713 | 64 | 18 | 92.16 | 93.31 | 91.76 | 79.68 | 85.96 | |
F-3 | 255 | 720 | 57 | 14 | 93.21 | 94.8 | 92.66 | 81.73 | 87.78 | |
F-4 | 255 | 720 | 57 | 14 | 93.21 | 94.8 | 92.66 | 81.73 | 87.78 | |
F-5 | 255 | 720 | 57 | 14 | 93.21 | 94.8 | 92.66 | 81.73 | 87.78 | |
Average | 92.68 | 94.05 | 92.2 | 80.68 | 86.86 | |||||
ResNet 50 | F-1 | 255 | 720 | 57 | 14 | 93.21 | 94.8 | 92.66 | 81.73 | 87.78 |
F-2 | 255 | 722 | 55 | 14 | 93.40 | 94.8 | 92.92 | 82.26 | 88.08 | |
F-3 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 | |
F-4 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 | |
F-5 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 | |
Average | 94.00 | 95.46 | 93.49 | 83.55 | 89.11 | |||||
Se-ResNet-50 | F-1 | 257 | 720 | 57 | 12 | 93.40 | 95.54 | 92.66 | 81.85 | 88.16 |
F-2 | 257 | 722 | 55 | 12 | 93.59 | 95.54 | 92.92 | 82.37 | 88.47 | |
F-3 | 259 | 732 | 45 | 10 | 94.74 | 96.28 | 94.21 | 85.2 | 90.4 | |
F-4 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 | |
F-5 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 | |
Average | 94.13 | 95.84 | 93.54 | 83.72 | 89.37 | |||||
Inception v4 | F-1 | 245 | 698 | 79 | 24 | 90.15 | 91.08 | 89.83 | 75.62 | 82.63 |
F-2 | 248 | 698 | 79 | 21 | 90.44 | 92.19 | 89.83 | 75.84 | 83.22 | |
F-3 | 248 | 704 | 75 | 21 | 90.84 | 92.19 | 90.37 | 76.78 | 83.78 | |
F-4 | 245 | 700 | 77 | 24 | 90.34 | 91.08 | 90.09 | 76.09 | 82.91 | |
F-5 | 245 | 698 | 79 | 24 | 90.15 | 91.08 | 89.83 | 75.62 | 82.63 | |
Average | 90.39 | 91.52 | 89.99 | 75.99 | 83.04 | |||||
Inception ResNet v2 | F-1 | 247 | 698 | 79 | 22 | 90.34 | 91.82 | 89.83 | 75.77 | 83.03 |
F-2 | 251 | 700 | 77 | 17 | 91.00 | 93.66 | 90.09 | 76.52 | 84.23 | |
F-3 | 251 | 714 | 63 | 17 | 92.34 | 93.66 | 91.89 | 79.94 | 86.25 | |
F-4 | 251 | 710 | 67 | 17 | 91.96 | 93.66 | 91.38 | 78.93 | 85.67 | |
F-5 | 247 | 698 | 79 | 22 | 90.34 | 91.82 | 89.83 | 75.77 | 83.03 | |
Average | 91.2 | 92.92 | 90.6 | 77.39 | 84.44 | |||||
ResNeXt-50 | F-1 | 258 | 730 | 47 | 11 | 94.46 | 95.91 | 93.95 | 84.59 | 89.9 |
F-2 | 260 | 732 | 45 | 9 | 94.84 | 96.65 | 94.21 | 85.25 | 90.59 | |
F-3 | 260 | 740 | 37 | 9 | 95.60 | 96.65 | 95.24 | 87.54 | 91.87 | |
F-4 | 260 | 740 | 37 | 9 | 95.60 | 96.65 | 95.24 | 87.54 | 91.87 | |
F-5 | 260 | 740 | 37 | 9 | 95.60 | 96.65 | 95.24 | 87.54 | 91.87 | |
Average | 95.22 | 96.51 | 94.77 | 86.49 | 91.22 | |||||
Se-ResNeXt-50 | F-1 | 260 | 732 | 45 | 9 | 94.84 | 96.65 | 94.21 | 85.25 | 90.59 |
F-2 | 261 | 733 | 44 | 8 | 95.03 | 97.03 | 94.34 | 85.57 | 90.94 | |
F-3 | 262 | 742 | 35 | 7 | 95.98 | 97.4 | 95.5 | 88.22 | 92.58 | |
F-4 | 262 | 742 | 35 | 7 | 95.98 | 97.4 | 95.5 | 88.22 | 92.58 | |
F-5 | 262 | 742 | 35 | 7 | 95.98 | 97.4 | 95.5 | 88.22 | 92.58 | |
Average | 95.56 | 97.17 | 95.01 | 87.09 | 91.85 |
Bold indicates highest accuracy