Table 3 ROC-AUC score as per feature reduction and supervised training approach
Model | Feature reduction | ROC-AUC |
|---|---|---|
Lasso as per [6] | – | 0.57 ± 0.05 |
\(DeepNet_i\) as per [6] | T-Test | 0.58 ± 0.04 |
Logistic regression | Lasso | 0.721 ± 0.059 |
PCA | 0.512 ± 0.029 | |
T-test | 0.724 ± 0.026 | |
Neural network | Lasso | 0.528 ± 0.091 |
PCA | 0.539 ± 0.116 | |
T-test | 0.657 ± 0.089 | |
Explainable boosting machines | Lasso | 0.581 ± 0.075 |
PCA | 0.534 ± 0.06 | |
T-test | 0.700 ± 0.056 | |
KNN | Lasso | 0.55 ± 0.09 |
PCA | 0.553 ± 0.089 | |
T-test | 0.544 ± 0.051 | |
XGBoost | Lasso | 0.581 ± 0.058 |
PCA | 0.579 ± 0.055 | |
T-test | 0.692 ± 0.061 |