Fig. 7
From: Cobdock: an accurate and practical machine learning-based consensus blind docking method
The feature importance for selected features by Boruta. Boruta selected features by optimising feature numbers. The selected features have been assessed using the ANOVA F-Test feature importance test. While the Y-axis demonstrates the F-score calculated by ANOVA, on the X-axis, the feature names in the format “PROGRAM_NAME”_“FEATURE_NAME” are displayed