Table 4 Performance of three classification decision tree model in training set and test set
Data Set | Indicators | Model One | Model Two | Model Three | |
|---|---|---|---|---|---|
Train Set | Accuracy (%) | 84.95 | 82.63 | 80.16 | |
Recall (%) | 24.17 | 86.78 | 83.47 | ||
Specificity (%) | 95.09 | 78.53 | 76.40 | ||
Youden index | 0.193 | 0.653 | 0.599 | ||
Kappa index | 0.192 | 0.653 | 0.575 | ||
PPV (%) | 49.13 | 79.97 | 64.46 | ||
NPV (%) | 86.47 | 85.74 | 90.01 | ||
Precision (%) | 98.85 | 78.53 | 80.75 | ||
F1 score (%) | 38.84 | 82.45 | 82.09 | ||
AUC (95%CI) | 0.627 (0.591,0.662) | 0.834 (0.805,0.861) | 0.832 (0.803,0.859) | ||
Test Set | Accuracy (%) | 84.35 | 82.75 | 78.59 | |
Recall (%) | 17.39 | 87.04 | 83.81 | ||
Specificity (%) | 93.63 | 78.15 | 75.48 | ||
Youden index | 0.110 | 0.652 | 0.593 | ||
Kappa index | 0.064 | 0.654 | 0.530 | ||
PPV (%) | 31.99 | 81.06 | 63.26 | ||
NPV (%) | 86.80 | 84.87 | 90.25 | ||
Precision (%) | 97.75 | 78.15 | 81.73 | ||
F1 score (%) | 29.53 | 82.36 | 82.76 | ||
AUC (95%CI) | 0.564 (0.507,0.619) | 0.824 (0.777,0.864) | 0.814 (0.766,0.855) | ||