Table 5 Assessment of the six prediction models.
Classification Models | AUC (95%CI) | Cutoff | Sensitivity | Specificity | FRP | TRP |
|---|---|---|---|---|---|---|
Training set | ||||||
GBM | 0.854(0.824–0.883) | 0.576 | 0.799 | 0.777 | 0.223 | 0.799 |
KNN | 0.863 (0.837–0.889) | 0.579 | 0.728 | 0.851 | 0.149 | 0.728 |
Light GBM | 0.9734(0.964–0.983) | 0.838 | 0.918 | 0.920 | 0.080 | 0.918 |
LR | 0.857 (0.829–0.885) | 0.580 | 0.774 | 0.806 | 0.194 | 0.774 |
NN | 0.853 (0.824–0.881) | 0.574 | 0.802 | 0.771 | 0.229 | 0.802 |
RF | 0.9867(0.981–0.993) | 0.877 | 0.940 | 0.937 | 0.063 | 0.940 |
Validation set | ||||||
GBM | 0.798 (0.744–0.852) | 0.493 | 0.850 | 0.643 | 0.357 | 0.850 |
KNN | 0.769 (0.709–0.830) | 0.443 | 0.757 | 0.686 | 0.314 | 0.757 |
Light GBM | 0.8034(0.748–0.859) | 0.521 | 0.749 | 0.771 | 0.229 | 0.749 |
LR | 0.797 (0.742–0.853) | 0.494 | 0.837 | 0.657 | 0.343 | 0.837 |
NN | 0.799(0.743–0.854) | 0.488 | 0.774 | 0.714 | 0.286 | 0.774 |
RF | 0.799(0.743–0.854) | 0.490 | 0.733 | 0.757 | 0.243 | 0.733 |
