Skip to main content

Table 2 Performance comparison of different models to predict 1-year relapse in the validation set

From: Multi-institutional development and external validation of machine learning-based models to predict relapse risk of pancreatic ductal adenocarcinoma after radical resection

Model AUC 95%CI.lower 95%CI.upper Sensitivity Specificity Accuracy PPV NPV F1 RMSE
LR 0.708 0.579 0.823 0.878 0.400 0.696 0.705 0.667 0.782 0.448
RF 0.653 0.519 0.782 0.837 0.400 0.671 0.695 0.600 0.759 0.501
SVM 0.733 0.603 0.840 0.857 0.467 0.709 0.724 0.667 0.785 0.445
GBM 0.560 0.416 0.708 0.776 0.367 0.620 0.667 0.500 0.717 0.509
NN 0.720 0.604 0.836 0.878 0.400 0.696 0.705 0.667 0.782 0.448
KNN 0.600 0.460 0.740 0.837 0.467 0.696 0.719 0.636 0.774 0.496