Fig. 5From: Intervention in prediction measure: a new approach to assessing variable importance for random forestsVIM barplots for the C-to-U conversion data. The first row corresponds to PVIM-CIT with mtry = 3, the second row to mtry = 43, while the bottom row to IPM-CIT-RF values with mtry = 43. In all cases, values are averaged over 100 replications. The variable names appear at bottom of the barplotsBack to article page