Table 3.
Mean (standard deviation) of performance metrics using the best 18 features selected by MRMR
Method | Accuracy | Recall | Precision | F1 | AUC |
---|---|---|---|---|---|
TFNN | 0.719 (0.033) | 0.617 (0.074) | 0.683 (0.046) | 0.600 (0.050) | 0.793 (0.039) |
XGB | 0.675 (0.027) | 0.584 (0.053) | 0.619 (0.033) | 0.554 (0.039) | 0.716 (0.034) |
RF | 0.740 (0.031) | 0.581 (0.053) | 0.744 (0.051) | 0.606 (0.044) | 0.800 (0.037) |
SVM | 0.731 (0.035) | 0.590 (0.055) | 0.719 (0.053) | 0.601 (0.047) | 0.800 (0.040) |
Values in bold are the highest for a given metric across different methods