NA, you
only calculate an aggregated value. If you can define a function that makes sense
for every single training / test set, implement your own Measure.makeCustomResampledMeasure(id, minimize = TRUE, properties = character(0L),
allowed.pred.types = character(0L), fun, extra.args = list(),
best = NULL, worst = NULL)Measure].G1, G2,
acc, auc, bac,
ber, cindex,
db, dunn, f1,
fdr, featperc,
fn, fnr, fp,
fpr, gmean,
gpr, mae, mcc,
mcp, meancosts,
measures, medae,
medse, mmce,
mse, multiclass.auc,
npv, ppv, rmse,
sae, silhouette,
sse, timeboth,
timepredict, timetrain,
tn, tnr, tp,
tpr; Measure,
makeMeasure; makeCostMeasure;
performance