Skip to main content

Table 3 Average over ten ten-fold cross-validation runs for given models trained on the 3-class classification subset

From: Selectivity profiling of BCRP versus P-gp inhibition: from automated collection of polypharmacology data to multi-label learning

Algorithm Embedded feature selection Accuracy (%) Kappa AUC
J48 tree Yes 61.6 0.42 0.73
J48 tree, min. 5 instances/leaf Yes 59.1 0.38 0.73
Jrip rules Yes 57.8 0.36 0.70
Bagging of J48 No 65.0 0.47 0.82
Bagging of J48 Yes 62.6 0.44 0.80