Table 8.

Five-fold cross-validation prediction performance of the proposed method and 4 state-of-the-art predictors on imbalanced benchmark dataset S₄ and S₅.

Methods	AUC	S4F1 score	AUPR	AUC	S5F1 score	AUPR
iMcRNA-PseSSC	0.9103	0.5707	0.6743	0.9333	0.7157	0.7628
iMiRNA-PseDPC	0.9333	0.6404	0.7317	0.9534	0.7708	0.8259
Triplet-SVM	0.8905	0.5207	0.6364	0.9357	0.7182	0.7806
miRNApre	0.9554	0.7029	0.7976	0.9454	0.7447	0.8140
Our method	0.9526	0.7084	0.7694	0.9589	0.7813	0.8525

The best values are shown in boldface.