Table 6.
Classification performance using PLV node features.
| Index | Repeat 1 | Repeat 2 | Repeat 3 | Repeat 4 | Repeat 5 | Mean ±SD | 95%CI |
|---|---|---|---|---|---|---|---|
| Accuracy (%) | 83.02 | 86.79 | 79.25 | 79.25 | 81.13 | 81.89 ± 3.15 | [78.16, 85.62] |
| Sensitivity (%) | 79.17 | 83.33 | 75.00 | 70.83 | 75.00 | 76.67 ± 4.75 | [70.80, 82.54] |
| Specificity (%) | 86.21 | 89.66 | 82.76 | 86.21 | 86.21 | 86.21 ± 2.44 | [83.15, 89.27] |
| Precision (%) | 82.61 | 86.96 | 78.26 | 80.95 | 81.82 | 82.12 ± 3.16 | [78.16, 86.08] |
| F1-score (%) | 80.85 | 85.11 | 76.60 | 75.56 | 78.38 | 79.30 ± 3.82 | [76.64, 83.96] |
| Balanced accuracy (%) | 82.69 | 86.49 | 78.88 | 78.52 | 80.60 | 81.44 ± 3.27 | [77.18, 85.70] |
| AUC (%) | 92.39 | 89.01 | 87.57 | 86.64 | 92.39 | 89.60 ± 2.68 | [86.43, 92.77] |