Table 4: Performance of various prediction algorithms across 5 folds of cross-validation.

ModelAverage AUCAUC standard
deviation

Linear-SVM 0.6910.016
RBF-SVM0.6900.008
Linear-SVM +
Class Cost
0.6950.014
Naïve Bayes0.6610.046
PLS-DA0.6600.029
LASSO0.7110.013
Random Forest0.6920.015
KNN0.6110.033
C4.50.6290.027

The bolded row shows the approach we have decided to use.