Pure interaction effects unseen by Random Forests

TL;DR

This paper investigates the limitations of Random Forests in capturing pure interaction effects and proposes alternative tree partitioning methods that improve interaction detection, validated through simulations and real data applications.

Contribution

It introduces modified tree-growing schemes that better identify pure interactions, addressing a gap in standard Random Forest algorithms.

Findings

Modified schemes outperform standard Random Forests in pure interaction scenarios

Enhanced detection of pure interactions in simulation studies

Improved model fitting on real datasets with interaction effects

Abstract

Random Forests are widely claimed to capture interactions well. However, some simple examples suggest that they perform poorly in the presence of certain pure interactions that the conventional CART criterion struggles to capture during tree construction. Motivated from this, it is argued that simple alternative partitioning schemes used in the tree growing procedure can enhance identification of these interactions. In a simulation study these variants are compared to conventional Random Forests and Extremely Randomized Trees. The results validate that the modifications considered enhance the model's fitting ability in scenarios where pure interactions play a crucial role. Finally, the methods are applied to real datasets.