Fig. 7

Variable importance from the CV run with the Random Forest classifier. Plot on the left (a) shows results for the main dataset and plot on the right (b) for the mystery-2 dataset. The order from top to bottom is given by the normalized score given by the classifier to each “species” at each split