Table 4 Comparison result from 20-time-10-fold cross validation with different sampling strategies
| Â | LR | CART | NB | cPDD |
|---|---|---|---|---|
Over Sampling | ||||
 F1-Score(T) | 0.31 ± 0.02 | 0.20 ± 0.02 | 0.26 ± 0.01 | 0.37 ± 0.01 |
 G-mean(T) | 0.34 ± 0.02 | 0.22 ± 0.22 | 0.36 ± 0.00 | 0.40 ± 0.01 |
 Avg. F1 | 0.31 ± 0.02 | 0.20 ± 0.20 | 0.26 ± 0.01 | 0.59 ± 0.01 |
 Balanced Acc. | 0.61 ± 0.01 | 0.52 ± 0.01 | 0.50 ± 0.00 | 0.61 ± 0.01 |
 MCC | 0.17 ± 0.03 | 0.04 ± 0.02 | 0.01 ± 0.00 | 0.25 ± 0.01 |
Under Sampling | ||||
 F1-Score(T) | 0.30 ± 0.01 | 0.27 ± 0.02 | 0.25 ± 0.01 | 0.34 ± 0.02 |
 G-mean(T) | 0.34 ± 0.01 | 0.30 ± 0.02 | 0.35 ± 0.02 | 0.41 ± 0.03 |
 Avg. F1 | 0.30 ± 0.01 | 0.27 ± 0.02 | 0.25 ± 0.01 | 0.63 ± 0.02 |
 Balanced Acc. | 0.59 ± 0.01 | 0.57 ± 0.01 | 0.54 ± 0.01 | 0.61 ± 0.02 |
 MCC | 0.13 ± 0.02 | 0.11 ± 0.02 | 0.08 ± 0.02 | 0.20 ± 0.08 |