Table 2 Evaluation metrics results of six models
Model | AUC (95%CI) | Accuracy (95%CI) | Sensitivity (95%CI) | Specificity (95%CI) | PPV (95%CI) | NPV (95%CI) | F1-score (95%CI) |
|---|---|---|---|---|---|---|---|
XGBoost | 0.782(0.715–0.849) | 0.763(0.724–0.802) | 0.633(0.518–0.749) | 0.824(0.720–0.927) | 0.309(0.267–0.352) | 0.936(0.932–0.940) | 0.41(0.372–0.447) |
LightGBM | 0.700(0.627–0.773) | 0.83(0.816–0.844) | 0.527(0.435–0.619) | 0.788(0.691–0.886) | 0.346(0.291–0.401) | 0.905(0.896–0.914) | 0.408(0.378–0.438) |
RF | 0.772(0.704–0.841) | 0.882(0.874–0.889) | 0.653(0.574–0.733) | 0.799(0.740–0.857) | 0.665(0.555–0.774) | 0.889(0.884–0.894) | 0.652(0.585–0.718) |
GBDT | 0.723(0.650–0.797) | 0.729(0.689–0.769) | 0.643(0.541–0.745) | 0.738(0.647–0.829) | 0.257(0.225–0.290) | 0.92(0.913–0.928) | 0.365(0.327–0.402) |
GNB | 0.562(0.483–0.642) | 0.47(0.253–0.688) | 0.747(0.520–0.973) | 0.416(0.174–0.657) | 0.155(0.142–0.168) | 0.918(0.884–0.953) | 0.254(0.223–0.286) |
SVM | 0.528(0.446–0.611) | 0.595(0.389–0.801) | 0.51(0.250–0.770) | 0.642(0.395–0.889) | 0.169(0.136–0.203) | 0.882(0.850–0.914) | 0.24(0.190–0.289) |