DeepAR: a novel deep learning-based hybrid framework for the interpretable prediction of androgen receptor antagonists

Table 2 Cross-validation results of different stacked models on the training dataset

Meta-model	ACC	Sn	Sp	MCC	AUC	F1
DeepAR	0.880	0.861	0.899	0.762	0.941	0.880
RF	0.884	0.887	0.881	0.770	0.952	0.887
LGBM	0.880	0.898	0.859	0.762	0.945	0.885
SVMRBF	0.878	0.893	0.862	0.758	0.923	0.882
XGB	0.877	0.901	0.850	0.755	0.942	0.882
LR	0.877	0.901	0.850	0.755	0.946	0.883
PLS	0.877	0.893	0.859	0.755	0.946	0.881
NB	0.877	0.887	0.866	0.755	0.919	0.881
ET	0.875	0.887	0.862	0.753	0.949	0.880
MLP	0.874	0.881	0.866	0.750	0.938	0.877
KNN	0.869	0.887	0.850	0.739	0.869	0.874
SVMLN	0.866	0.887	0.844	0.735	0.921	0.872
ADA	0.854	0.867	0.841	0.711	0.923	0.860
DT	0.853	0.846	0.859	0.709	0.853	0.854

ISSN: 1758-2946