Skip to main content

Table 2 Comparison of the predictive performance for six models (testing set)

From: The prediction of asymptomatic carotid atherosclerosis with electronic health records: a comparative study of six machine learning models

Model Acc (%) Sp (%) Pp (%) Re (%) F1 (%) AUCROC
LR 74.7 86.6 68.6 53.2 59.9 0.809
RF 74.5 89.5 71.3 47.2 56.8 0.794
DT 65.4 71.8 51.2 53.8 52.5 0.628
XGB 73.4 87.8 68.0 47.2 55.7 0.788
GNB 67.0 88.0 63.1 37.2 46.8 0.753
KNN 68.8 81.5 57.7 45.6 50.9 0.704
  1. Acc, accuracy; Sp, specificity; Pp, precision; Re, recall; F1, F1 score; AUCROC, the area under the receiver operating characteristic curve; LR, logistic regression; RF, random forest; DT, decision tree; XGB, eXtreme Gradient Boosting; GNB, Gaussian Naïve Bayes; KNN, K-Nearest Neighbour