Identification of Protein Pupylation Sites Using Bi-Profile Bayes Feature Extraction and Ensemble Learning

Table 3

The comparison of predictive performance between single SVM and ensemble of SVMs using the 5-fold cross validation on Dataset 1.

SVM classifier	Sensitivity (%)	Specificity (%)	Accuracy (%)	MCC	𝐴ROC
SinglePup	75.76	72.50	73.97	0.480	0.807
EnsemblePup	79.49	82.35	80.82	0.617	0.862