Performance of SVM classifiers.

Re-substitution test examines self-consistency of the classification method by classifying on the training set. Precision = TP/(TP+FP), Recall = TP/(TP+FN), F-measure = 2×(Precision×Recall)/(Precision+Recall), Accuracy = (TP+TN)/(TP+FP+TN+FN), where TP = true positive, TN = true negative, FP = false positive, and FN = false negative. The F-measure [81] is the harmonic mean of Precision and Recall, and is a particularly useful performance measure when the dataset is unbalanced such that there are significantly more negative examples than positive ones. We chose not to measure Specificity ( = TN/(TN+FP)) because it is less meaningful in such situations.