Table 3 Comparison of performance measures for all classifiers for the non-transactional dataset
From: Modelling customers credit card behaviour using bidirectional LSTM neural networks
Classifier | Accuracy (%) | Sensitivity (%) | Specificity (%) | Balanced accuracy (%) | AUC (%) | Brier score (%) | KS | H-measure |
|---|---|---|---|---|---|---|---|---|
GB | 82.07 | 36.47 | 95.02 | 65.75 | 78.03 | 13.43 | 0.43 | 0.29 |
BNN | 81.78 | 37.07 | 94.48 | 65.77 | 77.53 | 13.58 | 0.42 | 0.28 |
SVM | 81.46 | 28.56 | 96.49 | 62.52 | 69.57 | 14.33 | 0.37 | 0.23 |
RF | 80.18 | 17.01 | 98.12 | 57.57 | 76.54 | 14.38 | 0.41 | 0.27 |
LOGR | 80.88 | 22.56 | 97.44 | 60 | 71.82 | 14.55 | 0.37 | 0.24 |
Bidirectional LSTM | 82.4 | 37.51 | 95.15 | 66.33 | 78.47 | 13.28 | 0.43 | 0.3 |