An AUC-based permutation variable importance measure for random forests
An AUC-based permutation variable importance measure for random forests is a scholarly work by Anne-Laure Boulesteix, Silke Janitza, and Carolin Strobl, published in 2013 in ''BMC Bioinformatics''. The main subjects of the publication include mathematics, ranking function, computer science, cyclic permutation, artificial intelligence, receiver operating characteristic, resampling, Permutation, algorithm, data mining, class, random forest, machine learning, regularization, k-permutation, statistics, and random permutation. The authors investigated the performance of the standard permutation VIM and of the authors' novel AUC-based permutation VIM for different class imbalance levels using simulated data and real data.