Skip to main content

Table 1 Impact of the parameter λ on the performance of GTB classifer

From: Predicting effective drug combinations using gradient tree boosting based on features extracted from drug-protein heterogeneous network

λ Precision Recall F-Measure MCC AUC
0.1 0.861 0.852 0.856 0.715 0.929
0.2 0.865 0.853 0.858 0.720 0.930
0.3 0.870 0.854 0.861 0.726 0.934
0.4 0.878 0.861 0.869 0.738 0.939
0.5 0.883 0.871 0.877 0.755 0.941
0.6 0.885 0.868 0.875 0.755 0.941
0.7 0.887 0.867 0.876 0.757 0.944
0.8 0.885 0.865 0.875 0.754 0.943
0.9 0.884 0.864 0.874 0.752 0.943
  1. The boldface figures indicate that GTB classifier achieves the best performance at λ equal to 0.7