Skip to main content

Table 4 Comparison of AUC scores of four ML models using word vectors with different dimensions

From: Ontology-based venous thromboembolism risk assessment model developing from medical records

Dimension 10 20 40 80 100 120 150 200
AUC (GBDT) 0.863 ± 0.024 0.891 ± 0.023 0.916 ± 0.026 0.916 ± 0.017 0.929 ± 0.014 0.929 ± 0.015 0.927 ± 0.020 0.927 ± 0.020
AUC (RF) 0.852 ± 0.022 0.871 ± 0.030 0.883 ± 0.023 0.881 ± 0.029 0.884 ± 0.021 0.893 ± 0.020 0.897 ± 0.018 0.884 ± 0.019
AUC (LR) 0.851 ± 0.023 0.884 ± 0.019 0.912 ± 0.019 0.921 ± 0.020 0.926 ± 0.022 0.923 ± 0.022 0.920 ± 0.026 0.923 ± 0.014
AUC (SVM) 0.862 ± 0.019 0.873 ± 0.034 0.861 ± 0.031 0.879 ± 0.021 0.869 ± 0.025 0.880 ± 0.031 0.867 ± 0.023 0.830 ± 0.034
  1. The value of AUC score is formatted with ‘Mean value ± Standard deviation’. All terms are used to train models