TRIP detects unreliable feature importance scores in random forests.
problem Unreliable feature importance scores in random forests due to model extrapolation.
method Develops TRIP (Test for Reliable Interpretation via Permutation) to detect unreliable permutation feature importance scores.
result TRIP reliably detects unreliable permutation feature importance scores in high-dimensional settings.
New methods reduce extrapolation errors in feature importance.
problem Flawed feature importance methods using unrestricted permutations lead to extrapolation errors.
method Three new approaches: conditional model reliance, Knockoffs with Gaussian transformation, and restricted ALE plot designs.
result Theoretical and numerical results show our strategies reduce/eliminate extrapolation.
New method corrects correlation bias in feature importance.
problem Correlations between features in high-dimensional data bias statistical and machine learning models.
method Pairwise Permutation Algorithm (PPA)
result PPA corrects correlation bias, identifying biological relevant biomarkers.
A new method improves feature importance and model stress-testing reliability.
problem Estimating feature contributions in machine learning models for trust and transparency.
method Replacing multiple random permutations with a single, deterministic, and optimal permutation.
result Improved bias-variance tradeoffs and accuracy in challenging scenarios.
Two new methods assess feature importance for fairness in machine learning models.
problem Understanding how features influence fairness in machine learning models.
method Two model-agnostic approaches: permutation and occlusion.
result Simple, scalable, and interpretable methods to quantify feature importance for fairness.
TIME explains temporal models by analyzing feature importance.
problem Existing methods struggle with temporal models and feature importance.
method Model-agnostic permutation-based approach, temporal feature importance, hypothesis testing.
result TIME provides statistical rigor for explaining temporal models.
Introduces RFI for assessing feature importance relative to any subset of features.
problem Lack of nuanced feature importance computation.
method Generalizes PFI and CFI to assess relative feature importance.
result Derives general interpretation rules for RFI.
This paper evaluates and improves metrics for identifying important features in machine learning models.
problem Evaluation metrics for explainable AI are limited by multicollinearity and model accuracy.
method Proposes Expected Accuracy Interval (EAI) to predict model accuracy with multicollinearity.
result EAI is a useful metric for identifying important features in models with multicollinearity.
This study compares feature importance and explainability in quantum vs classical ML models.
problem Lack of transparency in ML models, especially in sensitive fields.
method Comparison of classical ML (SVM, Random Forest) and hybrid quantum ML (VQC, QSVC) models using feature importance and explainability methods.
result Quantum ML models provide insights similar to classical models but with unique quantum features.
Enhances functional classifier performance with new tree-based methods and unbiased feature importance assessment.
problem Challenges of high-dimensional functional data and biased feature importance assessment.
method Augmented functional classification trees and random forests with ad-hoc conditional permutations for unbiased feature importance.
result Significant enhancement in predictive power of functional classifiers through new feature importance assessment.
Electronic health records are an increasingly important resource for understanding the interactions between patient health, environment, and clinical decisions. In this paper we report an empirical study of predictive modeling of several patient outcomes using three state-of-the-art machine learning methods. Our primar…
The paper derives theoretical foundations for two common machine learning variable importance measures.
problem Understanding variable importance in machine learning problems.
method The paper derives closed-form expressions for Permute-and-Predict (PaP) and Leave-One-Covariate-Out (LOCO) methods.
result Theoretical derivations explain the behavior of PaP and LOCO under collinearity, linking them to coefficients and predictor variability.
Proposes a new method for interpreting feature importance and effects in dependent feature models.
problem Challenges in interpreting feature importance when features are dependent and interactions are present.
method Conditional Subgroup Approach
result Conditional PFI and PDP estimates based on this approach often outperform existing methods.
Researchers formalize PD and PFI to relate them to data generating process.
problem Lack of theory linking PD and PFI to data generating process.
method Formalize PD and PFI as estimators of ground truth estimands, account for model variance with learner-PD and learner-PFI.
result PD and PFI estimates deviate from ground truth due to statistical biases, model variance, and Monte Carlo approximation errors.
Study examines challenges in variable importance ranking due to feature correlation.
problem Challenges in variable importance ranking under correlation.
method Simulation study and theoretical analysis of feature knockoffs and conditional predictive impact (CPI).
result Highly correlated features increase the correlation of knockoff variables, posing a limitation for CPI.
Hollow-tree Super resolves feature importance in large datasets.
problem Lack of effective scaling for large feature numbers in boosted tree models.
method Hollow-tree Super (HOTS) methodology for feature importance visualization.
result HOTS effectively resolves feature importance and directionality in high-dimensional neuroscientific data.
Paper proves a Central Limit Theorem for Random Forest Permutation Importance Measure.
problem Lack of theoretical analysis of Random Forest Permutation Importance Measure (RFPIM).
method Formal proof using U-Statistics theory, deviating from conventional Random Forest model.
result Established a Central Limit Theorem for RFPIM.
Unordered feature sets are a nonstandard data structure that traditional neural networks are incapable of addressing in a principled manner. Providing a concatenation of features in an arbitrary order may lead to the learning of spurious patterns or biases that do not actually exist. Another complication is introduced …
Study evaluates feature ranking methods' faithfulness in ML models, improving with dimensionality reduction.
problem Quantifying and improving the faithfulness of feature ranking methods in ML models.
method Evaluation of multiple feature ranking methods, including SHAP, LIME, ALE variance, and LR coefficients, using permutation importance as a baseline.
result Dimensionality reduction improves the faithfulness of feature ranking methods, making permutation importance the most faithful method.
New tests detect high-order interactions without permutations.
problem Scalability issues in kernel-based tests for high-order interactions.
method Permutation-free high-order tests using V-statistics and cross-centring.
result Tests yield standard normal distribution under null hypothesis.
CPI overcomes limitations of permutation importance by providing accurate variable selection.
problem Misidentification of unimportant variables in complex models due to covariate correlations.
method Developed a model agnostic and computationally lean Conditional Permutation Importance (CPI) approach.
result CPI provides accurate type-I error control and more parsimonious variable selection.
Proposes a simple algorithm to generate data similar to real series.
problem Generating data similar to real series with constraints.
method Random permutation of Monte Carlo generated numbers, accepted if objective function is minimized.
result Demonstrated by generating S\&P 500 log-returns.
A2I Transformer predicts atom energies from coordinates, avoiding heavy featurization.
problem Efficiently predicting atom energies from molecular coordinates with minimal featurization.
method End-to-end model using self-attention, permutation-equivariant.
result Stable predictions with significantly smaller errors than molecular dynamics simulations.
Model-agnostic interpretation methods can mislead if not used carefully.
problem Misinterpretation of machine learning models due to improper use of techniques.
method General pitfalls of model-agnostic interpretation methods.
result Many pitfalls exist when using global interpretation techniques for machine learning models.
This paper reviews and advocates against the use of permute-and-predict (PaP) methods for interpreting black box functions. Methods such as the variable importance measures proposed for random forests, partial dependence plots, and individual conditional expectation plots remain popular because they are both model-agno…
New methods for assessing and visualizing feature groups in machine learning models.
problem Lack of methods for interpreting feature groups in machine learning models.
method Permutation-based, refitting, and Shapley-based techniques for grouped feature importance. Introduced a sequential procedure for identifying stable feature combinations. Developed a combined features effect plot.
result Effective methods for assessing and visualizing the importance and effect of feature groups in machine learning models.
Variable selection in sparse regression models is an important task as applications ranging from biomedical research to econometrics have shown. Especially for higher dimensional regression problems, for which the link function between response and covariates cannot be directly detected, the selection of informative va…
A new method reduces computational costs for testing RF variable importance measures.
problem Testing variable importance measures from random forests is computationally expensive and challenging.
method Sequential permutation testing and sequential p-value estimation to reduce computational costs.
result Theoretical properties of sequential tests are confirmed, maintaining type-I error and high power.
CIT and CIF improve feature selection for downstream prediction.
problem Feature selection bias in machine learning models.
method Conditional inference trees and forests with Bonferroni correction.
result CIF ranks top 3 among 18 regression methods and top 4 among 17 classification methods.
HKConv learns hyperbolic features by aggregating kernel points.
problem Challenges in learning good hyperbolic representations using Euclidean operations.
method Proposes HKConv, a trainable hyperbolic convolution that correlates local features with kernel points and aggregates them.
result HKConv learns expressive local features according to hyperbolic geometry and enjoys equivariance to permutation and invariance to parallel transport.
Training features used to analyse physical processes are often highly correlated and determining which ones are most important for the classification is a non-trivial tasks. For the use case of a search for a top-quark pair produced in association with a Higgs boson decaying to bottom-quarks at the LHC, we compare feat…
Study compares imputation methods' effects on IML confidence intervals.
problem Missing data impacts IML interpretation and confidence intervals.
method Compared single vs multiple imputation methods on IML confidence intervals.
result Multiple imputation provides closer coverage to nominal than single imputation.
XAI methods struggle with identifying true predictors from suppressors in linear datasets.
problem XAI methods misidentify suppressor variables as important features.
method Carefully crafted linear ground-truth dataset to study suppressor variables; evaluated various XAI methods.
result Most XAI methods fail to distinguish true predictors from suppressors in linear settings.
A new method resolves permutation issues in shuffled linear regression for large-scale applications.
problem Estimating latent features through linear transformation with unknown permutations.
method Spectral matching method to align spectral components of measurement and feature covariances.
result Achieves accurate estimates in shuffled LS and LASSO settings with sufficient samples.
New algorithm uses GNNs to optimize rewards in graph-structured data.
problem Optimizing rewards in molecule design with graph-structured data.
method Embedding permutation invariance into GNNs and using GNTK for regret bounds.
result First GNN confidence bound and phased-elimination algorithm with sublinear regret.
New methods interpret clustering outcomes without altering data structure.
problem Post-processing methods destroy data integrity and obscure interpretations.
method Algorithm-agnostic interpretation methods using permutation feature importance, individual conditional expectation, and partial dependence.
result Preserves original feature structure and explains clustering outcomes.
VarPro selects features without model dependence, achieving balanced performance.
problem Finding a small set of features with high explanatory power.
method Rule-based variable priority approach, avoiding model-specific methods and artificial data.
result VarPro has a consistent filtering property for noise variables and achieves balanced performance.
New approach interprets machine learning models through feature space transformations.
problem Interpreting models with strongly dependent features in high-dimensional spaces.
method Feature space transformations, including PCA and partial orthogonalization.
result Enhances model interpretation tools for domain experts.
AI techniques explain synthetic tabular data weaknesses.
problem Challenges in evaluating synthetic tabular data quality.
method Apply explainable AI to a binary detection classifier.
result Reveals inconsistencies, unrealistic dependencies, or missing patterns in synthetic data.
ExNODE uses ODE to model sets with permutation equivariance.
problem Capturing intra-set dependencies in unordered sets.
method Exchangeable Neural ODE (ExNODE) using ODE.
result ExNODE achieves permutation equivariance for set modeling.
Proposes CLIQUE for improved local variable importance in multi-class classification.
problem Lack of methods to characterize local structure in model loss space.
method CLIQUE (Conditional Local Importance by Quantile Expectations)
result CLIQUE emphasizes locally dependent information and captures interaction behavior.
New graph foundation models respect symmetries for broader applicability.
problem Tailored graph machine learning architectures limit broader applicability.
method Investigates symmetries for label and feature permutations, proving network universal approximator.
result Universal approximator on multisets respecting node and feature permutations.
Cheap permutation tests speed up distribution testing without sacrificing accuracy.
problem Efficiently testing distribution differences and independence.
method Group datapoints into bins and permute only these bins, using stored sufficient statistics.
result Cheap permutation tests maintain the accuracy and optimality of standard tests but are significantly faster.
Permutation-valued features arise in a variety of applications, either in a direct way when preferences are elicited over a collection of items, or an indirect way in which numerical ratings are converted to a ranking. To date, there has been relatively limited study of regression, classification, and testing problems …
Improves modeling of sets with permutation invariant densities.
problem Challenges in calculating trace limit practicality of current methods.
method Proposes an alternative approach to define permutation equivariant transformations with closed form trace.
result Improves both training and final performance.
Bayesian optimization method for permutations accelerates combinatorial search.
problem Optimizing expensive-to-evaluate objectives on permutation problems.
method LAW2ORDER, a batch Bayesian optimization method based on the acquisition weighted kernel.
result LAW2ORDER achieves sublinear batch cumulative regret, demonstrating accelerated search.
The paper extends explainability methods to uncertainty-aware models, revealing feature impacts on predictive entropy and likelihood.
problem Understanding the factors contributing to uncertainty in probabilistic models.
method Adapting permutation feature importance, partial dependence plots, and individual conditional expectation plots to measure feature impacts on predictive entropy and likelihood.
result Novel insights into model behaviour and feature impacts on uncertainty are obtained.
The paper examines properties of GW optimal transport plans, showing they can be sparse and permutation-supported.
problem Properties of Gromov-Wasserstein optimal transport plans.
method Exploration of sparsity, permutation support, and cyclical monotonicity properties.
result GW optimal plans can be sparse and permutation-supported under certain conditions.