Table 2.

Specific Performance of Four Methods on Dataset RPI488.

RPI488	Accuracy	Sensitivity	Specificity	Precision	MCC
BGFE	0.8868	0.9268	0.8354	0.9328	0.7743
Raw feature	0.8168	0.8083	0.8192	0.8104	0.6299
Stacked auto-encoder	0.8806	0.9243	0.8255	0.9351	0.7638
Stacked auto-encoder without fine tuning	0.8600	0.8848	0.8271	0.8850	0.7187

The boldface indicates this measure performance is the best among the compared methods for individual dataset.