Title: Variable importance in clustering using binary decision trees
Authors: Pierre Michel - Aix Marseille University (AMU) (France) [presenting]
Badih Ghattas - Aix Marseille University (France)
Abstract: Different approaches are considered for assessing variable importance in clustering. We focus on clustering using binary decision trees, which is a non-parametric top-down hierarchical clustering method designed for both continuous and nominal data. We suggest a measure of variable importance for this method similar to the one used in Breiman's classiffcation and regression trees. We analyze the effciency of this score on different data simulation models in presence of noise, and compare it to other classical variable importance measures.