mlsanityData to 14 Oct 2026

Unbiased Measurement of Feature Importance in Tree-Based Methods

arXiv 1903.05179 ↗First posted 12 Mar 2019stat.ML cs.LGcited by 30

Not on the map. Why?

Abstract (arXiv)

We propose a modification that corrects for split-improvement variable importance measures in Random Forests and other tree-based methods. These methods have been shown to be biased towards increasing the importance of features with more potential splits. We show that by appropriately incorporating split-improvement as measured on out of sample data, this bias can be corrected yielding better summaries and screening tools.

Key facts. Cited by 30 papers in the map (data through 14 Oct 2026). First posted on arXiv 12 Mar 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 8; 2022: 4; 2023: 5; 2024: 5; 2025: 3; 2026: 5.

30 papers.

Sort:

  1. 2021S-LIME: Stabilized-LIME for Model Explanationcited by 52
  2. 2022Fast Interpretable Greedy-Tree Sumscited by 13
  3. 2021MDA for random forests: inconsistency, and a practical solution via the Sobol-MDAcited by 10
  4. 2021ControlBurn: Feature Selection by Sparse Forestscited by 9
  5. 2021Feature Importance Explanations for Temporal Black-Box Modelscited by 7
  6. 2021Bridging Breiman's Brook: From Algorithmic Modeling to Statistical Learningcited by 7
  7. 2025GlyTwin: Digital Twin for Glucose Control in Type 1 Diabetes Through Optimal Behavioral Modifications Using Patient-Centric Counterfactualscited by 5
  8. 2022Individualized and Global Feature Attributions for Gradient Boosted Trees in the Presence of ℓ2 Regularizationcited by 2
  9. 2023Unbiased Gradient Boosting Decision Tree with Unbiased Feature Importancecited by 2
  10. 2021Modeling Household Online Shopping Demand in the U.S.: A Machine Learning Approach and Comparative Investigation between 2009 and 2017cited by 1
  11. 2021A Subspace-based Approach for Dimensionality Reduction and Important Variable Selectioncited by 1
  12. 2022ControlBurn: Nonlinear Feature Selection with Sparse Tree Ensemblescited by 1
18 more
  1. 2023Interpreting Deep Forest through Feature Contribution and MDI Feature Importancecited by 1
  2. 2024McUDI: Model-Centric Unsupervised Degradation Indicator for Failure Prediction AIOps Solutionscited by 1
  3. 2025Tackling Small Sample Survival Analysis via Transfer Learning: A Study of Colorectal Cancer Prognosiscited by 1
  4. 2021Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient databasecited by 0
  5. 2022A Novel Splitting Criterion Inspired by Geometric Mean Metric Learning for Decision Treecited by 0
  6. 2023The Berkelmans-Pries Feature Importance Method: A Generic Measure of Informativeness of Featurescited by 0
  7. 2023CIMLA: Interpretable AI for inference of differential causal networkscited by 0
  8. 2023End-to-end Feature Selection Approach for Learning Skinny Treescited by 0
  9. 2024Evaluating Explanatory Capabilities of Machine Learning Models in Medical Diagnostics: A Human-in-the-Loop Approachcited by 0
  10. 2024Predicting the duration of traffic incidents for Sydney greater metropolitan area using machine learning methodscited by 0
  11. 2024Multi forests: Variable importance for multi-class outcomescited by 0
  12. 2024Iterative Feature Exclusion Ranking for Deep Tabular Learningcited by 0
  13. 2025Provable Recovery of Locally Important Signed Features and Interactions from Random Forestcited by 0
  14. 2026Towards Public Administration Research Based on Interpretable Machine Learningcited by 0
  15. 2026Unity Forests: Improving Interaction Modelling and Interpretability in Random Forestscited by 0
  16. 2026SeqShield: A Behavioral Analysis Approach to Uncover Rootkitscited by 0
  17. 2026Just add noise: Debiasing tree-based variable importance in mixed datacited by 0
  18. 2026Splitting the Difference: Interpretable Causal Forests for Treatment Effect Heterogeneity and Biascited by 0