Computing local-level variable importance (VI) is critical for understanding machine learning models. Local permutation-based VI metrics offer a simple yet powerful approach, measuring how much a mode...
arXiv:2511.15371v3 Announce Type: replace Abstract: Assessing the importance of individual features in Machine Learning is critical to understand the model's decision-making process. While numerous m...
arXiv:2512. 13892v3 Announce Type: replace-cross Abstract: Reliable estimation of feature contributions in machine learning models is essential for transparency, algorithmic fairness, and regulatory compliance.
arXiv:2609.36396v1 Announce Type: cross Abstract: As black-box machine learning models become increasingly common, extracting interpretations with uncertainty quantification has become a critical cha...
arXiv:2510.12734v2 Announce Type: replace Abstract: Variable importance (VI) methods are often used for hypothesis generation, feature selection, and scientific validation. In the standard VI pipelin...
arXiv:2603.05575v2 Announce Type: replace-cross Abstract: We study prediction-powered conditional inference in the setting where labeled data are scarce, unlabeled covariates are abundant, and a blac...
arXiv:2604. 15107v2 Announce Type: replace-cross Abstract: Shapley values provide a flexible framework for attributing feature contributions to model predictions, but they are not naturally suited for feature selection: a feature may receive a positive attribution even when it is redundant given the remaining variables.
arXiv:2411. 08821v4 Announce Type: replace-cross Abstract: Global variable importance measures are commonly used to interpret the results of machine learning models.
The paper introduces PRIME, a new variable‑importance measure that addresses two limitations of the existing proportional marginal effects (PME) method: it is applicable to stochastic models and reduces computational cost. PRIME achieves this by employing a model‑free variable‑selection technique to isolate active variables before computing importance. Experiments on synthetic test functions and real datasets demonstrate that PRIME accurately and reliably quantifies variable importance from noisy input‑output data.
The paper introduces SHAP concentration as a pre‑deployment diagnostic for detecting when conformal prediction will fail under distribution shift, specifically in gradient‑boosted classifiers. Using a COVID‑19 supply‑chain case study, the authors show that higher feature‑importance concentration correlates with larger drops in coverage, while standard shift detectors cannot differentiate between catastrophic and robust outcomes. The diagnostic is validated on additional datasets, and a formal theorem links concentration to worsening conformity‑score bounds, though it does not capture global‑sensitivity failures in neural networks.
arXiv:2512. 13003v2 Announce Type: replace-cross Abstract: Out-of-distribution (OOD) detection is essential for determining when a supervised model encounters inputs that differ meaningfully from its training distribution.
The paper introduces MCIR-M, a global feature‑importance method that accounts for feature dependence by conditioning each feature on its strongly dependent neighbours and computing a normalized Mutual Correlation Impact Ratio. MCIR-M yields a population score between 0 and 1, reaching zero when a feature is exactly conditionally redundant, and can be estimated with only a fraction of the data. Experiments on synthetic redundancy setups and the UCI HAR benchmark show that MCIR-M produces more stable rankings under near‑duplicate predictors compared to existing methods such as SHAP, SAGE, HSIC, and MI‑based scores.