Random forest regression analysis showing the relative importance of activity and attention measures in classifying subjects as ADHD or controls. Variables are rank ordered by importance, which was determined in two ways. The left panel indicates importance by how much the permutation (effective elimination) of a given variable decreases the accuracy of the overall fit. The right panel indicates importance by how much the permutation of a variable attenuates the ability of the specific nodes in the random forest to accurately split the sample. Variables that are associated with the greatest decrease in accuracy or Gini coefficient following permutation are the most important.