From: Synthesizing class labels for highly imbalanced credit card fraud detection data
Classifier | Df | Sum Sq | Mean SQ | F value | Pr(>F) | Df | Sum Sq | Mean SQ |
---|---|---|---|---|---|---|---|---|
Positive labels | Residuals | |||||||
DT | 20 | 1.0176 | 0.05088 | 63.52 | < 2e−16 | 1029 | 0.8242 | 0.0008 |
RF | 20 | 0.2861 | 0.014306 | 7.771 | < 2e−16 | 1029 | 1.8943 | 0.001841 |
ET | 20 | 0.1707 | 0.008534 | 4.782 | 2.99E−11 | 1029 | 1.8362 | 0.001784 |
LR | 20 | 0.8621 | 0.0431 | 24.44 | < 2e−16 | 1029 | 1.8149 | 0.00176 |
NB | 20 | 0.0288 | 0.001441 | 3.251 | 1.96E−06 | 1029 | 0.4561 | 0.000443 |
MLP | 20 | 0.9528 | 0.04764 | 27.41 | < 2e−16 | 1029 | 1.7883 | 0.00174 |