Feature Importance Measurement based on Decision Tree Sampling

TL;DR

This paper introduces DT-Sampler, a SAT-based method for measuring feature importance in decision tree models, offering improved interpretability and stability over traditional random forest approaches.

Contribution

The paper presents a novel SAT-based feature importance measurement method that requires fewer parameters and enhances interpretability and stability in real-world applications.

Findings

DT-Sampler provides higher interpretability than random forest.

The method demonstrates increased stability in feature importance analysis.

Implementation is publicly available at GitHub.

Abstract

Random forest is effective for prediction tasks but the randomness of tree generation hinders interpretability in feature importance analysis. To address this, we proposed DT-Sampler, a SAT-based method for measuring feature importance in tree-based model. Our method has fewer parameters than random forest and provides higher interpretability and stability for the analysis in real-world problems. An implementation of DT-Sampler is available at https://github.com/tsudalab/DT-sampler.