Table 5 Model performance based on several evaluation metrics.
Non-landslide sampling method | Model | Sensitivity | Specificity | Precision | Accuracy | F1-score |
|---|---|---|---|---|---|---|
PU bagging | RF | 0.696 | 0.786 | 0.765 | 0.741 | 0.729 |
SVM | 0.768 | 0.839 | 0.827 | 0.804 | 0.796 | |
CatBoost | 0.839 | 0.804 | 0.810 | 0.821 | 0.825 | |
BCS | RF | 0.750 | 0.643 | 0.677 | 0.696 | 0.712 |
SVM | 0.750 | 0.696 | 0.712 | 0.723 | 0.730 | |
CatBoost | 0.696 | 0.679 | 0.684 | 0.688 | 0.690 | |
KM | RF | 0.929 | 1.000 | 1.000 | 0.964 | 0.963 |
SVM | 0.929 | 0.982 | 0.981 | 0.955 | 0.954 | |
CatBoost | 0.946 | 0.946 | 0.946 | 0.946 | 0.946 |
