Logo Logo
Switch Language to German
Hornung, Roman; Boulesteix, Anne-Laure (9. March 2021): Interaction Forests: Identifying and exploiting interpretable quantitative and qualitative interaction effects. Department of Statistics: Technical Reports, No.237


Although interaction effects can be exploited to improve predictions and allow for valuable insights into covariate interplay, they are given little attention in analysis. We introduce interaction forests, which are a variant of random forests for categorical, continuous, and survival outcomes, explicitly considering quantitative and qualitative interaction effects in bivariable splits performed by the trees constituting the forests. The new effect importance measure (EIM) associated with interaction forests allows ranking of the covariate pairs with respect to their interaction effects' importance for prediction. Using EIM, separate importance value lists for univariable effects, quantitative interaction effects, and qualitative interaction effects are obtained. In the spirit of interpretable machine learning, the bivariable split types of interaction forests target well interpretable interaction effects that are easy to communicate. To learn about the nature of the interplay between identified interacting covariate pairs it is convenient to visualise their estimated bivariable influence. We provide functions that perform this task in the R package diversityForest that implements interaction forests. In a large-scale empirical study using 220 data sets, interaction forests tended to deliver better predictions than conventional random forests and competing random forest variants that use multivariable splitting. In a simulation study, EIM delivered considerably better rankings for the relevant quantitative and qualitative interaction effects than competing approaches. These results indicate that interaction forests are suitable tools for the challenging task of identifying and making use of well interpretable interaction effects in predictive modelling.