Skip to main content

Table 2 Area measures and c statistics are shown for 3 parts of an ROC curve i = {1 . . . 3} as well as the whole curve, for a classifier, a support vector machine, applied to Ljubljana breast cancer remission data. Best values per column are shown in bold font

From: A new concordant partial AUC and partial c statistic for imbalanced data in the evaluation of machine learning algorithms

i

FPR

TPR

pAUC

pAUCc

pAUCx

c∆

1

[0.00, 0.33]

[0.00, 0.84]

21.3%

49.5%

77.7%

49.5%

2

[0.33, 0.66]

[0.84, 0.95]

29.5%

17.4%

5.3%

17.4%

3

[0.66, 1.00]

[0.95, 1.00]

34.0%

17.9%

1.8%

17.9%

sum

–

–

84.8%

84.8%

84.8%

84.8%

whole

AUC = c = 84.8%AUPRC+,− = 72.2, 53.7%