Table 3 Diagnosis efficiency of different machine learning algorithms in training groups
AUC | Sensitivity | Specificity | |
|---|---|---|---|
Adaboost | 0.9973(0.9962,0.9975,0.9986) | 0.9748(0.9647,0.9762,0.9878) | 0.9804(0.9750,0.9875,0.9880) |
Xgboost | 0.9945(0.9911,0.9954,0.9969) | 0.9612(0.9529,0.9643,0.9647) | 0.9747(0.9747,0.9756,0.9759) |
SVM | 0.9935(0.9915,0.9928,0.9955) | 0.9632(0.9529,0.9639,0.9643) | 0.9625(0.9524,0.9639,0.9643) |
RF | 0.9975(0.9972,0.9987,0.9994) | 0.9855(0.9762,0.9880,1.0000) | 0.9860(0.9765,0.9880,1.0000) |
logistic | 0.9880(0.9847,0.9867,0.9904) | 0.9522(0.9398,0.9512,0.9634) | 0.9457(0.9398,0.9412,0.9524) |
naivebayes | 0.9732(0.9692,0.9720,0.9765) | 0.9222(0.9143,0.9178,0.9296) | 0.8174(0.7959,0.8061,0.8280) |
NPV | PPV | MCC | |
Adaboost | 0.9744(0.9639,0.9759,0.9880) | 0.9803(0.9759,0.9880,0.9881) | 0.9549(0.9402,0.9524,0.9759) |
Xgboost | 0.9605(0.9518,0.9639,0.9643) | 0.9750(0.9759,0.9759,0.9762) | 0.9357(0.9280,0.9398,0.9407) |
SVM | 0.9632(0.9524,0.9639,0.9643) | 0.9624(0.9518,0.9639,0.9643) | 0.9257(0.9157,0.9277,0.9398) |
RF | 0.9853(0.9759,0.9880,1.0000) | 0.9858(0.9759,0.9880,1.0000) | 0.9713(0.9639,0.9759,0.9880) |
logistic | 0.9524(0.9398,0.9518,0.9639) | 0.9453(0.9398,0.9405,0.9518) | 0.8978(0.8797,0.8923,0.9159) |
naivebayes | 0.9336(0.9277,0.9286,0.9398) | 0.7900(0.7590,0.7738,0.8072) | 0.7316(0.7003,0.7184,0.7445) |