Skip to main content

Table 3 SNF-NN performance comparison with baseline machine learning methods on the Cdataset 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 0.666 0.683 0.672 0.649 0.660 0.333 0.666 0.748
DT 0.611 0.809 0.711 0.412 0.505 0.253 0.610 0.709
GPC 0.707 0.604 0.692 0.811 0.738 0.433 0.707 0.799
KNN 0.695 0.560 0.654 0.830 0.731 0.406 0.695 0.785
GNB 0.654 0.645 0.651 0.662 0.656 0.308 0.654 0.741
QDA 0.629 0.472 0.599 0.786 0.679 0.272 0.629 0.746
RF 0.618 0.612 0.618 0.624 0.620 0.237 0.618 0.715
Linear-SVM 0.692 0.673 0.685 0.712 0.698 0.386 0.692 0.771
RBF-SVM 0.530 1.000 0.994 0.060 0.112 0.172 0.530 0.762
SNF-NN 0.783 0.754 0.769 0.813 0.790 0.569 0.879 0.856
  1. The best value of each evaluation metric is shown in bold