Skip to main content

Table 4 SNF-NN performance comparison with baseline machine learning methods on the LRSSL benchmark dataset

From: SNF-NN: computational method to predict drug-disease interactions using similarity network fusion and neural networks

Method Accuracy Specificity Precision Recall F1-score MCC AUC-ROC AUC-PR
SAMME 0.665 0.705 0.680 0.624 0.650 0.331 0.665 0.746
DT 0.635 0.775 0.689 0.494 0.574 0.282 0.635 0.718
GPC 0.701 0.717 0.709 0.685 0.696 0.403 0.701 0.776
KNN 0.661 0.553 0.633 0.769 0.694 0.329 0.661 0.759
GNB 0.616 0.509 0.596 0.723 0.653 0.238 0.616 0.729
QDA 0.567 0.662 0.613 0.473 0.453 0.161 0.567 0.675
RF 0.611 0.683 0.631 0.539 0.580 0.225 0.611 0.700
Linear-SVM 0.678 0.672 0.676 0.685 0.680 0.357 0.678 0.759
RBF-SVM 0.578 0.869 0.687 0.286 0.403 0.191 0.578 0.665
SNF-NN 0.846 0.780 0.821 0.793 0.807 0.617 0.936 0.903
  1. The best value of each evaluation metric is shown in bold