Skip to main content
. 2022 Feb 22;13:790063. doi: 10.3389/fmicb.2022.790063

TABLE 2.

Performance of descriptors before and after feature selection and in feature fusion.

Descriptors SN SP AAC MCC AUC
Before feature selection ACC 0.9304 0.9308 0.9306 0.8626 0.9723
tPseAAC 0.9011 0.8793 0.8899 0.7914 0.9551
aPseAAC 0.8901 0.8720 0.8808 0.7714 0.9519
DC 0.7546 0.8720 0.8149 0.5963 0.8812
DDE 0.8022 0.8374 0.8203 0.6319 0.9081
CKSAAP 0.7912 0.5398 0.6619 0.3855 0.7365
CTD 0.9377 0.9100 0.9235 0.8612 0.9786
After feature selection ACC 0.9524 0.9239 0.9377 0.8902 0.9735
tPseAAC 0.8938 0.8962 0.8950 0.7943 0.9580
aPseAAC 0.8971 0.8824 0.8895 0.7863 0.9610
DC 0.8859 0.8754 0.8416 0.6620 0.9143
DDE 0.7802 0.8651 0.8238 0.6430 0.9165
CKSAAP 0.7070 0.8374 0.7740 0.5156 0.8349
CTD 0.9167 0.9135 0.9150 0.8330 0.9644
Feature fusion 0.9634 0.9619 0.9626 0.9269 0.9864