Skip to main content
. 2021 Aug 23;21(16):5657. doi: 10.3390/s21165657

Table 1.

Performance metrics of the selected ensemble as well as of each of its 4 individual models.

Accuracy F1Score AUC
Model Train Val Test Train Val Test Train Val Test
VGGval831 0.871 0.831 0.764 0.871 0.831 0.764 0.948 0.906 0.867
Effval856 0.995 0.856 0.788 0.995 0.856 0.788 1.000 0.884 0.828
Denseval898 0.988 0.898 0.803 0.988 0.898 0.803 1.000 0.963 0.880
Denseval907 1.000 0.907 0.803 1.000 0.907 0.803 1.000 0.917 0.830
Ensemble 1.000 0.966 0.833 1.000 0.966 0.833 1.000 0.963 0.894