Table 2 The comparison of the three model performance
Selection approach | Feature size | Cohort | AUC (95%CI) | Accuracy | Sensitivity | Specificity |
|---|---|---|---|---|---|---|
LASSO24 | 24 | Training | 0.851(0.764–0.928) | 0.778 | 0.759 | 0.787 |
Validation | 0.782(0.585–0.954) | 0.816 | 0.667 | 0.885 | ||
SVM19 | 19 | Training | 0.845(0.748–0.934) | 0.811 | 0.862 | 0.787 |
Validation | 0.779(0.586–0.955) | 0.816 | 0.667 | 0.885 | ||
Random Forest20 | 20 | Trianing | 1.000(1.000–1.000) | 1.000 | 1.000 | 1.000 |
Validation | 0.780(0.608–0.926) | 0.790 | 0.583 | 0.885 |