Table 1 Comparison between the proposed model-based SMOTE and RUS-based model using different classifiers
From: A hybrid Stacking-SMOTE model for optimizing the prediction of autistic genes
Classifiers | Evaluation metrics | RUS% | SMOTE-RUS% | SMOTE % |
|---|---|---|---|---|
NB | Accuracy | 74.8 | 68.6 | 76.1 |
Precision | 77.9 | 71.8 | 83.6 | |
Recall | 74.9 | 68.6 | 76.0 | |
F-measure | 76.2 | 69.8 | 79.0 | |
RF | Accuracy | 84.2 | 86.3 | 90.7 |
Precision | 79.7 | 86.4 | 90.0 | |
Recall | 84.2 | 86.3 | 90.7 | |
F-measure | 79.5 | 85.3 | 88.8 | |
SVM | Accuracy | 71.5 | 81.9 | 88.4 |
Precision | 71.2 | 82.7 | 90.2 | |
Recall | 71.5 | 82.0 | 88.4 | |
F-measure | 71.3 | 82.3 | 89.1 | |
KNN | Accuracy | 79.4 | 87.9 | 92.2 |
Precision | 79.8 | 90.1 | 94.1 | |
Recall | 79.4 | 87.9 | 92.5 | |
F-measure | 79.6 | 88.4 | 93.0 |