Skip to main content

Table 7 Comparison of classification performance of machine learning methods on validating dataset of overall nodules

From: Diagnosis of thyroid nodules for ultrasonographic characteristics indicative of malignancy using random forest

  AUC SEN F1 SPE PPV NPV
RF 0.798 0.822 0.588 0.672 0.896 0.522
LR 0.820 0.810 0.611 0.729 0.912 0.526
SVM 0.781 0.777 0.558 0.685 0.895 0.470
NET 0.818 0.789 0.601 0.743 0.914 0.505
ELM 0.811 0.806 0.573 0.671 0.894 0.500
KNN 0.738 0.707 0.492 0.657 0.877 0.393
NB 0.796 0.851 0.587 0.629 0.888 0.550
ADAB 0.748 0.777 0.506 0.599 0.870 0.437
LOG 0.811 0.810 0.610 0.729 0.912 0.526
LDA 0.820 0.793 0.588 0.714 0.906 0.500
\