Skip to main content

Advertisement

Table 3 Publc validation set performance for all models and descriptor sets

From: Feature combination networks for the interpretation of statistical machine learning models: application to Ames mutagenicity

  MACCS Pubchem CDK standard
  AUC BAC SEN SPEC AUC BAC SEN SPEC AUC BAC SEN SPEC
SVM 0.87 0.81 0.82 0.79 0.88 0.82 0.84 0.80 0.87 0.81 0.84 0.78
RF 0.88 0.82 0.86 0.77 0.88 0.81 0.86 0.76 0.88 0.81 0.83 0.79
DT 0.81 0.77 0.80 0.74 0.79 0.76 0.78 0.74 0.80 0.75 0.79 0.72
kNN 0.84 0.76 0.84 0.68 0.84 0.77 0.81 0.73 0.83 0.75 0.81 0.70
  CDK Extended Atom centered     
  AUC BAC SEN SPEC AUC BAC SEN SPEC     
SVM 0.87 0.81 0.83 0.79 0.88 0.82 0.84 0.80     
RF 0.87 0.80 0.82 0.78 0.88 0.81 0.82 0.80     
DT 0.78 0.75 0.80 0.71 0.79 0.75 0.79 0.71     
kNN 0.84 0.77 0.81 0.73 0.84 0.77 0.82 0.72     
  1. AUC = area under curve, BAC = balanced accuracy, SEN = sensitivity, SPEC = specificity.