Table 4.

Averaged test set accuracy of ML models, where $F_{ALL}$ is the number of all features, and $F_S$ means the number of features selected using the RFE algorithm.

Model	${\mathit{F}}_{\mathbf{ALL}}$ = 1444	${\mathit{F}}_{\mathit{S}}$ = 1000	${\mathit{F}}_{\mathit{S}}$ = 722	${\mathit{F}}_{\mathit{S}}$ = 361	${\mathit{F}}_{\mathit{S}}$ = 160	${\mathit{F}}_{\mathit{S}}$ = 100
RF	0.8294	0.8280	0.8256	0.8250	0.8284	0.8258
SVM	0.7850	0.7920	0.7964	0.8126	0.7931	0.7695
XGB	0.8318	0.8283	0.8342	0.8287	0.8230	0.8177
LR	0.7795	0.7828	0.7952	0.8111	0.7910	0.7682
ANN	0.8223	0.8269	0.8210	0.8283	0.8185	0.8100