Learning Decision Trees and Forests with Algorithmic Recourse

TL;DR

This paper introduces a novel algorithm for learning accurate decision trees and forests that guarantees the existence of actionable recourse for most instances, balancing predictive accuracy with actionable fairness.

Contribution

It proposes a new greedy algorithm leveraging adversarial training to ensure recourse actions while maintaining high accuracy, applicable to both trees and forests.

Findings

Provides more instances with feasible recourse actions than baselines

Maintains comparable predictive accuracy to traditional methods

Operates efficiently on large datasets

Abstract

This paper proposes a new algorithm for learning accurate tree-based models while ensuring the existence of recourse actions. Algorithmic Recourse (AR) aims to provide a recourse action for altering the undesired prediction result given by a model. Typical AR methods provide a reasonable action by solving an optimization task of minimizing the required effort among executable actions. In practice, however, such actions do not always exist for models optimized only for predictive performance. To alleviate this issue, we formulate the task of learning an accurate classification tree under the constraint of ensuring the existence of reasonable actions for as many instances as possible. Then, we propose an efficient top-down greedy algorithm by leveraging the adversarial training techniques. We also show that our proposed algorithm can be applied to the random forest, which is known as a popular framework for learning tree ensembles. Experimental results demonstrated that our method successfully provided reasonable actions to more instances than the baselines without significantly degrading accuracy and computational efficiency.