Table 4 The models’ performance when dealing with imbalanced data across both sweep and fixed frequencies files.
Dataset | Models | F1_score | POD | FAR | TSS | HSS |
|---|---|---|---|---|---|---|
Sweep frequency (imbalance) | dtree | 0.68(±0.04) | 0.76(±0.04) | 0.36(±0.09) | 0.67(±0.02) | 0.62(±0.05) |
|  | RF | 0.75(±0.03) | 0.85(±0.08) | 0.30(±0.05) | 0.78(±0.07) | 0.71(±0.03) |
|  | svm | 0.70(±0.01) | 0.76(±0.08) | 0.33(±0.08) | 0.68(±0.05) | 0.64(±0.02) |
|  | linsvm | 0.68(±0.06) | 0.78(±0.12) | 0.38(±0.04) | 0.70(±0.11) | 0.62(±0.07) |
Fixed frequency (imbalance) | dtree | 0.66(±0.08) | 0.70(±0.12) | 0.36(±0.08) | 0.65(±0.11) | 0.62(±0.09) |
|  | RF | 0.71(±0.06) | 0.76(±0.12) | 0.31(±0.08) | 0.71(±0.11) | 0.67(±0.06) |
|  | svm | 0.7(±0.04) | 0.82(±0.07) | 0.37(±0.09) | 0.75(±0.05) | 0.65(±0.05) |
|  | linsvm | 0.70(±0.06) | 0.85(±0.06) | 0.38(±0.09) | 0.78(±0.05) | 0.66(±0.07) |
