# Consistent feature attribution for tree ensembles

arXiv: https://arxiv.org/abs/1706.06060 · HTML: https://mlsanity.com/paper/1706.06060/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jun 2017.

## Built on by (most cited first, 18 of 18)

- [Synthetic Benchmarks for Scientific Research in Explainable Machine Learning](https://mlsanity.com/paper/2106.12543/) (2021)
- [Interpreting Attributions and Interactions of Adversarial Attacks](https://mlsanity.com/paper/2108.06895/) (2021)
- [Explainable AI for Bioinformatics: Methods, Tools, and Applications](https://mlsanity.com/paper/2212.13261/) (2022)
- [Coalitional strategies for efficient individual prediction explanation](https://mlsanity.com/paper/2104.00765/) (2021)
- [Interpreting Black-box Machine Learning Models for High Dimensional Datasets](https://mlsanity.com/paper/2208.13405/) (2022)
- [RUMBoost: Gradient Boosted Random Utility Models](https://mlsanity.com/paper/2401.11954/) (2024)
- [Understanding the Dependence of Perception Model Competency on Regions in an Image](https://mlsanity.com/paper/2407.10543/) (2024)
- [Your Face Mirrors Your Deepest Beliefs-Predicting Personality and Morals through Facial Emotion Recognition](https://mlsanity.com/paper/2112.12455/) (2021)
- [Feature construction using explanations of individual predictions](https://mlsanity.com/paper/2301.09631/) (2023)
- [Action Shapley: A Training Data Selection Metric for World Model in Reinforcement Learning](https://mlsanity.com/paper/2601.10905/) (2026)
- [Early prediction of respiratory failure in the intensive care unit](https://mlsanity.com/paper/2105.05728/) (2021)
- [Multi-objective optimization and explanation for stroke risk assessment in Shanxi province](https://mlsanity.com/paper/2107.14060/) (2021)
- [Enhancing operations management through smart sensors: measuring and improving well-being, interaction and performance of logistics workers](https://mlsanity.com/paper/2112.08213/) (2021)
- [Validation of a Hospital Digital Twin with Machine Learning](https://mlsanity.com/paper/2303.04117/) (2023)
- [Designing Explainable Predictive Machine Learning Artifacts: Methodology and Practical Demonstration](https://mlsanity.com/paper/2306.11771/) (2023)
- [Are We Merely Justifying Results ex Post Facto? Quantifying Explanatory Inversion in Post-Hoc Model Explanations](https://mlsanity.com/paper/2504.08919/) (2025)
- [ParamBoost: Gradient Boosted Piecewise Cubic Polynomials](https://mlsanity.com/paper/2604.18864/) (2026)
- [LOCO-AdaMP: Built-in LOCO Inference for Adaptive Minipatch Ensembles with Enhanced Prediction](https://mlsanity.com/paper/2609.36396/) (2026)

## Abstract (arXiv)

Note that a newer expanded version of this paper is now available at: arXiv:1802.03888 It is critical in many applications to understand what features are important for a model, and why individual predictions were made. For tree ensemble methods these questions are usually answered by attributing importance values to input features, either globally or for a single prediction. Here we show that current feature attribution methods are inconsistent, which means changing the model to rely more on a given feature can actually decrease the importance assigned to that feature. To address this problem we develop fast exact solutions for SHAP (SHapley Additive exPlanation) values, which were recently shown to be the unique additive feature attribution method based on conditional expectations that is both consistent and locally accurate. We integrate these improvements into the latest version of XGBoost, demonstrate the inconsistencies of current methods, and show how using SHAP values results in significantly improved supervised clustering performance. Feature importance values are a key part of understanding widely used models such as gradient boosting trees and random forests, so improvements to them have broad practical implications.