Table 5 Performances base on Test-10 dataset using different thresholds. In bold, three models achieve high efficiency, especially BLF that reaches SPE near 90% with SEN 85%.
SEN % THRES | SPE % | SPE % THRES | SEN % | ||||
---|---|---|---|---|---|---|---|
ORI | BON | BLF | ORI | BON | BLF | ||
100.0 | 18.9 | 10.3 | 2.3 | 100.0 | 18.9 | 20.4 | 11.8 |
95.0 | 68.1 | 69.3 | 61.3 | 95.0 | 70.0 | 74.4 | 77.8 |
90.0 | 81.4 | 80.5 | 79.9 | 90.0 | 83.8 | 82.4 | 84.9 |
85.0 | 88.6 | 88.8 | 89.9 | 85.0 | 87.3 | 87.3 | 88.4 |
80.0 | 92.0 | 91.8 | 93.6 | 80.0 | 90.4 | 90.2 | 89.8 |
76.7 | 93.3 | 94.1 | 95.8 | 69.2 | 94.2 | 95.1 | 93.1 |
74.1 | 93.7 | 95.2 | 96.7 | 60.0 | 96.7 | 96.0 | 95.6 |