Table 3 Performance metrics of the models on the training and test set.
Models | AUC(95% CI) | Accuracy(95% CI) | Precision (95% CI) | Recall (95% CI) | F1 Score(95% CI) |
|---|---|---|---|---|---|
LR Test | 0.858 (0.8213, 0.8919) | 0.73 (0.6997, 0.7593) | 0.974 (0.9603, 0.986) | 0.703 (0.6706, 0.738) | 0.816 (0.7938, 0.8391) |
LR Train | 0.874 (0.8165, 0.9234) | 0.758 (0.7063, 0.8104) | 0.961 (0.9308, 0.9839) | 0.75 (0.6936, 0.8034) | 0.843 (0.8039, 0.8787) |
NNET Test | 0.866 (0.8045, 0.9232) | 0.881 (0.8401, 0.9182) | 0.891 (0.8498, 0.9263) | 0.983 (0.9646, 0.9958) | 0.934 (0.9098, 0.9558) |
NNET Train | 0.986 (0.9999, 1.0) | 0.893 (0.8722, 0.9143) | 0.983 (0.9723, 0.9913) | 0.902 (0.8795, 0.9235) | 0.94 (0.9275, 0.9526) |
RF Test | 0.844 (0.7722, 0.9069) | 0.862 (0.8215, 0.9033) | 0.865 (0.8246, 0.9063) | 0.996 (0.9863, 1.0) | 0.926 (0.902, 0.949) |
RF Train | 0.986 (0.95, 0.993) | 0.867 (0.8436, 0.8895) | 0.998 (0.9956, 1.0) | 0.867 (0.8427, 0.8894) | 0.928 (0.914, 0.941) |
SVM Test | 0.87 (0.8124, 0.9188) | 0.818 (0.7732, 0.8661) | 0.946 (0.9132, 0.9754) | 0.836 (0.7885, 0.8844) | 0.888 (0.8564, 0.9199) |
SVM Train | 0.986 (0.9999, 1.0) | 0.769 (0.7406, 0.799) | 0.765 (0.7353, 0.7973) | 0.956 (0.9388, 0.9722) | 0.85 (0.8299, 0.8718) |
XGBoost Test | 0.896 (0.8416, 0.941) | 0.896 (0.855, 0.9293) | 0.898 (0.8588, 0.9328) | 0.991 (0.9781, 1.0) | 0.943 (0.9195, 0.9622) |
XGBoost Train | 0.986 (0.9999, 1.0) | 0.909 (0.8895, 0.928) | 0.993 (0.9854, 0.9985) | 0.91 (0.8888, 0.93) | 0.949 (0.9378, 0.9604) |
Stacking Test | 0.911 (0.8653, 0.9494) | 0.903 (0.8661, 0.9368) | 0.912 (0.877, 0.9444) | 0.983 (0.9646, 0.9957) | 0.946 (0.9237, 0.965) |
Stacking Train | 0.986 (0.9999, 1.0) | 0.916 (0.8945, 0.9342) | 0.988 (0.9797, 0.9956) | 0.919 (0.8981, 0.9388) | 0.952 (0.9406, 0.9633) |
