mlsanityData to 7 Oct 2026

Targeted Learning for Variable Importance

arXiv 2411.02221 ↗First posted 4 Nov 2024stat.ML cs.LG stat.MEcited by 0

Not on the map. Why?

Abstract (arXiv)

Variable importance is one of the most widely used measures for interpreting machine learning with significant interest from both statistics and machine learning communities. Recently, increasing attention has been directed toward uncertainty quantification in these metrics. Current approaches largely rely on one-step procedures, which, while asymptotically efficient, can present higher sensitivity and instability in finite sample settings. To address these limitations, we propose a novel method by employing the targeted learning (TL) framework, designed to enhance robustness in inference for variable importance metrics. Our approach is particularly suited for conditional permutation variable importance. We show that it (i) retains the asymptotic efficiency of traditional methods, (ii) maintains comparable computational complexity, and (iii) delivers improved accuracy, especially in finite sample contexts. We further support these findings with numerical experiments that illustrate the practical advantages of our method and validate the theoretical results.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 22 of its 49 references that are in the map. First posted on arXiv 4 Nov 2024.

Builds on

22 of 49 references.

References · 22

  1. 2016Distribution-Free Predictive Inference For Regressioncited by 657
  2. 2018All Models are Wrong, but Many are Useful: Learning a Variable's Importance by Studying an Entire Class of Prediction Models Simultaneouslycited by 364
  3. 2022Semiparametric doubly robust targeted double machine learning: a reviewcited by 129
  4. 2019Unrestricted Permutation forces Extrapolation: Variable Importance Requires at least One More Model, or There Is No Free Variable Importancecited by 85
  5. 2018Deep Neural Networks for Estimation and Inferencecited by 62
  6. 2020Efficient nonparametric statistical inference on population feature importance using Shapley valuescited by 61
  7. 2020A general framework for inference on algorithm-agnostic variable importancecited by 44
  8. 2023The Rashomon Importance Distribution: Getting RID of Unstable, Single Model-based Variable Importancecited by 29
  9. 2014A General Theory of Hypothesis Tests and Confidence Regions for Sparse High Dimensional Modelscited by 26
  10. 2021Relating the Partial Dependence Plot and Permutation Feature Importance to the Data Generating Processcited by 24
  11. 2023Feature Importance: A Closer Look at Shapley Values and LOCOcited by 23
  12. 2022Scientific Inference With Interpretable Machine Learning: Analyzing Models to Learn About Real-World Phenomenacited by 21
10 more
  1. 2021Decorrelated Variable Importancecited by 12
  2. 2016High-Dimensional L2Boosting: Rate of Convergencecited by 10
  3. 2020Getting Better from Worse: Augmented Bagging and a Cautionary Tale of Variable Importancecited by 10
  4. 2019A Unified Framework for Random Forest Prediction Error Estimationcited by 8
  5. 2023Statistically Valid Variable Importance Assessment through Conditional Permutationscited by 7
  6. 2022LOCO Feature Importance Inference without Data Splitting via Minipatch Ensemblescited by 5
  7. 2023Kernel Debiased Plug-in Estimation: Simultaneous, Automated Debiasing without Influence Functions for Many Target Parameterscited by 5
  8. 2017Machine Learning and the Future of Realismcited by 4
  9. 2022Efficient Targeted Learning of Heterogeneous Treatment Effects for Multiple Subgroupscited by 3
  10. 2023Application of targeted maximum likelihood estimation in public health and epidemiological studies: a systematic reviewcited by 3

Built on by

Not cited yet.