Table 2.

Averaged AUC in cross-study validation setting.

Approach	No FS			FS only			FS + AE			DeepGeni (FS + DBG + AE)
Algorithm	SVM	RF	NN	SVM	RF	NN	SVM	RF	NN	SVM	RF	NN
AUC	0.52	0.522	0.556	0.564	0.551	0.585	0.602	0.57	0.598	0.626	0.579	0.609
STD	0.156	0.074	0.07	0.107	0.103	0.08	0.06	0.053	0.045	0.209	0.09	0.221

The bolded values indicate the best performance for each approach.

FS feature selection, AE autoencoder, DBG DeepBioGen.