A Two-Step Feature Selection Method to Predict Cancerlectins by Multiview Features and Synthetic Minority Oversampling Technique

Table 2

Performance comparisons of different machine learning methods on the full features using 5-fold cross-validation.

Machine learning method	Sensitivity	Specificity	Accuracy	MCC
AdaBoost	0.690	0.540	0.615	0.233
Decision Table	0.681	0.540	0.611	0.223
Nearest Neighbor Analysis	0.757	0.584	0.670	0.346
Logistic Regression	0.531	0.558	0.544	0.089
Naïve Bayes	0.500	0.699	0.600	0.203
RBFNetwork	0.615	0.491	0.553	0.107
Random Forest	0.704	0.695	0.699	0.398