Figure 3From: Prediction of beta-turns at over 80% accuracy based on an ensemble of predicted secondary structures and multiple alignments Selection of the desired number of features based on 3 fold cross-validation test on the BT426-20 dataset using SVM with C = 3 and γ = 0.0186; the x -axis shows the number of selected top features, while y -axis shows the corresponding accuracy (Q total ) of β -turn prediction. Back to article page