A0979
Title: Multi forests: Variable importance for multi-class outcomes
Authors: Roman Hornung - University of Munich (Germany) [presenting]
Alexander Hapfelmeier - Technical University of Munich (Germany)
Abstract: In prediction tasks with multi-class outcomes, identifying covariates specifically associated with one or more outcome classes can be important. Conventional variable importance measures (Vims) from random forests (Rfs), like permutation and Gini importance, focus on overall predictive performance or node purity without differentiating between the classes. Therefore, they can be expected to fail to distinguish class-associated covariates. A new Vim called multi-class Vim is tailored to identify exclusively class-associated covariates via a novel Rf variant called multi forests (Mufs). The trees in Mufs use both multi-way and binary splitting. The multi-way splits generate child nodes for each class, using a split criterion that evaluates how well these nodes represent their respective classes. This setup forms the basis of the multi-class Vim, which measures the discriminatory ability of the splits performed in the respective covariates with regard to this split criterion. Simulation studies demonstrate that the multi-class Vim specifically ranks class-associated covariates highly, unlike conventional Vims, which also rank other types of covariates highly. Analyses of 121 datasets reveal that Mufs often have slightly lower predictive performance compared to conventional Rfs. This is, however, not a limiting factor given the algorithm's primary goal of calculating the multi-class Vim.
