Skip to main content

Table 2 Predictive performance of various models in the training, test and external validation sets

From: Multi-algorithms analysis for pre-treatment prediction of response to transarterial chemoembolization in hepatocellular carcinoma on multiphase MRI

Classifiers

AUC

ACC

Sensitivity

Specificity

Training set

 KNN

0.774

0.704

0.5

0.898

 SVM

0.871

0.765

0.891

0.695

 Lasso

0.941

0.861

0.982

0.780

 DNN

0.927

0.870

0.911

0.864

Test set

 KNN

0.669

0.655

0.538

0.75

 SVM

0.688

0.621

0.769

0.563

 Lasso

0.745

0.655

0.769

0.813

 DNN

0.837

0.759

0.923

0.688

External validation set

 KNN

0.615

0.536

0.857

0.357

 SVM

0.712

0.679

0.786

0.714

 Lasso

0.663

0.679

0.929

0.500

 DNN

0.796

0.714

0.714

0.857

  1. KNN, k-nearest neighbor; SVM, support vector machine; Lasso, the least absolute shrinkage and selection operator; DNN, deep neural networks; AUC, the area under the receiver operating characteristic curve; ACC, accuracy. Bold represents the highest values of AUC, ACC, sensitivity, and specificity in different data sets