Table 5 Comparison of evaluation scores of different models.
From: Site selection and prediction of urban emergency shelter based on VGAE-RF model
Compare objects | Precision | Recall | F1 score | Imbalance index |
|---|---|---|---|---|
Existing shelters | – | – | – | 0.4436 |
RF | 0.8623 | 0.8467 | 0.8492 | 0.4642 |
SVM | 0.8211 | 0.8604 | 0.8344 | 0.4667 |
KNN | 0.8267 | 0.8493 | 0.8378 | 0.4635 |
MLP | 0.8673 | 0.8950 | 0.8809 | 0.4582 |
VGAE-SVM | 0.8528 | 0.8995 | 0.8756 | 0.4426 |
VGAE-KNN | 0.8405 | 0.8904 | 0.8647 | 0.4463 |
VGAE-MLP | 0.8722 | 0.9041 | 0.8879 | 0.4421 |
VGAE-RF-NF | 0.8616 | 0.8813 | 0.8713 | 0.4573 |
VGAE-RF | 0.9112 | 0.8904 | 0.9018 | 0.4403 |
