Table 8 Results of the test dataset for locoregional recurrence
Particular | Data | ||
|---|---|---|---|
Model name | locoregional Recurrence | ||
No. of samples | 311 | 311 | 311 |
Total number of features | 384 | 384 | 384 |
ML algorithm used | Random Forest | KSVM | XgBoost |
Feature selection technique | SFFS | SFFS | SFFS |
OverSampling method used | SMOTE | SMOTE | SMOTE |
No. of samples after OverSampling | 514 | 514 | 514 |
Number of synthetic samples | 203 | 203 | 203 |
No. of features used in the dataset | 216 | 24 | 169 |
Mean accuracy train | 0.96 | 0.97 | 1 |
Mean accuracy test | 0.66 | 0.77 | 0.62 |
Sensitivity | 0.75 | 0.95 | 1.00 |
Specificity | 0.63 | 0.89 | 0.62 |
Mean F-Score train label 0 | 0.96 | 0.97 | 1 |
Mean F-Score train label 1 | 0.96 | 0.97 | 1 |
Mean F-score test label 0 | 0.78 | 0.82 | 0.76 |
Mean F-score test label 1 | 0.26 | 0.68 | 0.07 |
No. of samples for class-0 | 257 | 257 | 257 |
No. of samples for class-1 | 54 | 54 | 54 |
Base algorithm | Random forest | KSVM | XGBoost |
ROC_AUC_Score | 0.89 | 0.94 | 0.84 |