Intervention in prediction measure: a new approach to assessing variable importance for random forests
Random forests are a popular method in many fields since they can be successfully applied to complex data, with a small sample size, complex interactions and correlations, mixed type predictors, etc. Furthermo...
BMC Bioinformatics 2017 18:230
Published on: 2 May 2017