Table 16.
# of synthetic samples | ||||||||
---|---|---|---|---|---|---|---|---|
5k | 10k | 20k | ∼170k (all) | |||||
Precision | Recall | Precision | Recall | Precision | Recall | Precision | Recall | |
IM | 0.488 | 0.599 | 0.496 | 0.680 | 0.496 | 0.745 | 0.497 | 0.867 |
BN | 0.497 | 0.863 | 0.496 | 0.918 | 0.497 | 0.918 | 0.498 | 0.966 |
MPoM | 0.492 | 0.826 | 0.491 | 0.864 | 0.492 | 0.891 | 0.497 | 0.945 |
CLGP | 0.492 | 0.880 | 0.493 | 0.903 | 0.498 | 0.937 | 0.500 | 0.981 |
MC-MedGAN | 1.000 | 0.002 | 1.000 | 0.002 | 1.000 | 0.002 | 1.000 | 0.002 |
MICE-LR | 0.491 | 0.842 | 0.497 | 0.884 | 0.493 | 0.912 | 0.498 | 0.964 |
MICE-LR-DESC | 0.493 | 0.809 | 0.496 | 0.857 | 0.492 | 0.884 | 0.496 | 0.955 |
MICE-DT | 0.496 | 0.899 | 0.498 | 0.936 | 0.499 | 0.955 | 0.502 | 0.996 |
BREAST large-set. Values shown are the average performance over 10 independently generated synthetic samples