New method R-LOCO improves local feature importance analysis.
problem Local attribution methods fail to accurately identify important features.
method R-LOCO segments input space into regions and applies global methods within.
result R-LOCO delivers more accurate local attributions.
New local MDI variable importances derived from global scores match Shapley values.
problem Local feature relevance in tree-based models.
method Deriving local MDI importance measure from global scores and linking it to Shapley values.
result Local MDI importances have a natural connection with Shapley values.
New method disentangles feature importance scores in machine learning.
problem Misinterpretation of feature importance scores due to interactions and dependencies.
method Derive DIP (Disentangled Importance) decomposition of feature importance scores.
result DIP decomposition uniquely separates standalone contributions from interactions and dependencies.
SAGE quantifies feature importance in machine learning models.
problem Understanding the role of individual features in complex models.
method Formalizing predictive power through model-based and universal measures, and introducing SAGE for efficient calculation.
result SAGE assigns more accurate feature importance values than other methods.
In recent years, a large amount of model-agnostic methods to improve the transparency, trustability and interpretability of machine learning models have been developed. We introduce local feature importance as a local version of a recent model-agnostic global feature importance method. Based on local feature importance…
DEDACT breaks down feature importance into direct and associative components.
problem Lack of clear distinction between direct and associative feature importance.
method DEDACT framework to decompose direct and associative importance measures.
result Provides insight into sources of prediction-relevant information and feature pathways.
Global Sensitivity Analysis improves feature importance ranking in Random Forests.
problem Improving feature importance ranking in Random Forests.
method Applying Global Sensitivity Analysis to Random Forests for feature ranking.
result Our method provides a novel way to rank features based on their importance.
Introduces joint Shapley values to measure feature importance in models.
problem Measuring the importance of feature sets in machine learning models.
method Extends Shapley's axioms to measure a set of features' average contribution to a model's prediction.
result Joint Shapley values provide unique insights and are more consistent with local intuitions.
A new method averages neural network parameters to rank features robustly.
problem Neural networks' sensitivity to random initialization affects feature ranking robustness.
method Parameter averaging of multiple shallow networks trained with different random seeds.
result The averaged model discovers ground-truth feature importance consistently.
The problem of explaining deep learning models, and model predictions generally, has attracted intensive interest recently. Many successful approaches forgo global approximations in order to provide more faithful local interpretations of the model's behavior. LIME develops multiple interpretable models, each approximat…
New method visualizes tabular feature semantics for better model understanding.
problem Lack of feature interaction interpretation in tabular ML models.
method Feature Vectors method for global tabular dataset interpretability.
result Visualizes semantic relationships among tabular features.
A new method ranks and selects features without model fitting.
problem Feature importance measures algorithm-specific and need improvement.
method Integrates global sensitivity analysis with forward selection and backward elimination.
result Demonstrates clear advantage over state-of-the-art methods.
New methods compare local and global feature importance scores for bioinformatics models.
problem Improving feature importance estimation in tree-based models.
method Comparison of SHAP values and Conditional Feature Contributions (CFCs) for 164 bioinformatics problems.
result SHAP values and CFCs yield similar rankings and interpretations for random forests.
Local MDI+ improves feature importance for tree-based models, enhancing interpretability and performance.
problem Lack of instance-specific feature importance for tree-based models.
method Local MDI+ extends MDI+ framework to provide instance-specific feature importances.
result Local MDI+ outperforms existing baselines, improving predictive performance by 10%.
Proposes φ-table for statistical SHAP explanations in regression models.
problem Lack of clear directional summaries, uncertainty, and fidelity in SHAP feature importance.
method SHAP importance selection, fitting a standardized linear surrogate, reporting coefficients, uncertainty, fidelity, and stability.
result Extends SHAP into a statistical global explanation with direction, uncertainty, fidelity, and stability.
A framework for quantifying uncertainty in feature importance values.
problem Stable interpretation of feature importance values in machine learning models.
method A novel method based on pairwise comparisons of feature importance values to produce confidence intervals for feature ranks.
result The method produces simultaneous confidence intervals for feature ranks, enabling selection of top-k important features.
New method identifies important features and interactions in RF models.
problem Limited theoretical understanding of local feature and interaction importance in RF models.
method Combines global and local analysis to identify frequent feature co-occurrences.
result Proves consistent recovery of true local signal features and interactions.
A framework infers feature importance with uncertainties for high-dimensional data.
problem Estimating feature importance in high-dimensional data with uncertainty.
method Shapley value based framework, sub-SAGE, bootstrapping.
result Uncertainties in feature importance can be estimated from bootstrapping.
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.
Efficiently estimates SAGE values using causal structure learning.
problem Computational infeasibility of exact SAGE calculations.
method Uses causal structure learning to identify conditional independencies and accelerate SAGE approximation.
result Empirically demonstrates efficient and accurate estimation of SAGE values.
Quantifying the importance and power of individual nodes depending on their position in socio-economic networks constitutes a problem across a variety of applications. Examples include the reach of individuals in (online) social networks, the importance of individual banks or loans in financial networks, the relevance …
New methods ensure feature importance rankings are correct with high probability.
problem Stability issues in feature importance scores due to random sampling.
method Hypothesis testing-based techniques to assess and verify the stability of top-ranked features.
result Ensures the most important features are correct with high-probability guarantees.
Feature selection methods are widely used in order to solve the 'curse of dimensionality' problem. Many proposed feature selection frameworks, treat all data points equally; neglecting their different representation power and importance. In this paper, we propose an unsupervised hypergraph feature selection method via …
BIF assesses feature importance using Dirichlet distribution and Bayesian inference.
problem Quantitative feature importance assessment in statistical models.
method Utilizes Dirichlet distribution for probabilistic feature importance assessment via approximate Bayesian inference.
result Learned importance provides relative significance and confidence quantification of features.
Understanding the behavior of a trained network and finding explanations for its outputs is important for improving the network's performance and generalization ability, and for ensuring trust in automated systems. Several approaches have previously been proposed to identify and visualize the most important features by…
GOLFS selects features for clustering by combining global and local information.
problem Feature selection for high-dimensional clustering without labels.
method Combines global and local information via manifold learning and regularized self-representation.
result Improves feature selection and clustering accuracy.
The paper introduces closed-form expressions for interpreting Tsetlin Machines.
problem Interpreting complex Tsetlin Machines with a large number of clauses.
method Developed closed-form expressions for local and global interpretability of Tsetlin Machines.
result The expressions enable real-time feature importance assessment and data clustering.
MACQ method explains deep learning models by analyzing feature contributions across prediction levels.
problem Explaining deep learning model predictions.
method Global gradient-based, model-agnostic approach focusing on marginal attribution.
result MACQ separates feature contributions from interaction effects and visualizes 3-way relationships.
A barrier to the wider adoption of neural networks is their lack of interpretability. While local explanation methods exist for one prediction, most global attributions still reduce neural network decisions to a single set of features. In response, we present an approach for generating global attributions called GAM, w…
The fragmentation of production across countries has become an important feature of the globalization in recent decades and is often conceptualized by the term, global value chains (GVCs). When empirically investigating the GVCs, previous studies are mainly interested in knowing how global the GVCs are rather than how …
This paper improves land cover classification using global spatial features in CNN.
problem Limited classification accuracy and universality of traditional remote sensing image classification methods.
method Integrates global spatial features into a dual-branch CNN for hyperspectral/SAR imagery classification.
result The proposed method outperforms traditional single-channel CNN methods.
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 i…
Magnitude-based features capture interactions between different entities in multispecies spatial data.
problem Capturing interactions between different entities in multispecies spatial data.
method Developing magnitude-based features for multispecies spatial data.
result Identifies distinct neighbourhood types and spatial heterogeneity.
Proposes ICE-based metric for better understanding interactions in black-box models.
problem Misleading global sensitivity metrics in black-box models due to interaction effects.
method Individual Conditional Expectation (ICE) curves to compute feature importance and interactions.
result ICE-based metric provides richer insights into feature importance and interactions.
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 provides a guide to feature importance methods for better scientific inference.
problem Limited understanding of data-generating process due to opaque ML model mechanisms.
method Comprehensive review and new proofs of global feature importance methods.
result Facilitates a thorough understanding and concrete recommendations for FI methods.
New FI method accurately predicts feature importance.
problem Difficulty in evaluating feature importance in machine learning models.
method Combines Shapley values and Berkelmans-Pries dependency function.
result Proves accurate FI values for various cases.
Improved local feature attributions using neighbourhood reference distributions.
problem Misleading results from global population in local model behaviour.
method Formulation of neighbourhood reference distributions and self-normalised importance sampling.
result Neighbourhood Shapley values provide meaningful sparse feature attributions.
A new method uses asymmetric Shapley values to assess gene importance in clinical prediction models.
problem Clinical prediction models struggle with assessing the importance of high-dimensional features like genomics.
method Derive efficient algorithms to compute local and global asymmetric Shapley values for a mixed-dimensional prediction model.
result Asymmetric Shapley values provide a more suitable alternative to quantify feature importance in clinical prediction models.
The paper explores how Shapley value for a feature can vary based on model outcomes and feature distribution.
problem The uniqueness of Shapley value in explaining model predictions.
method Analyzes the relationship between feature distribution and Shapley value, and compares Shapley values for different model outcomes.
result Shapley value for a feature depends on more than just its mean and can vary significantly based on model outcome.
While the success of deep neural networks (DNNs) is well-established across a variety of domains, our ability to explain and interpret these methods is limited. Unlike previously proposed local methods which try to explain particular classification decisions, we focus on global interpretability and ask a universally ap…
The paper develops statistical inference methods for SHAP values.
problem Lack of statistical inference for SHAP values in model-agnostic feature importance.
method Semi-parametric approach using U-statistics and Neyman orthogonal scores for functionals of nested regressions.
result Asymptotically normal estimates of the pth powers of SHAP values for various p.
There is intense interest in understanding the stochastic and dynamical properties of the global Foreign Exchange (FX) market, whose daily transactions exceed one trillion US dollars. This is a formidable task since the FX market is characterized by a web of fluctuating exchange rates, with subtle inter-dependencies wh…
The goal of feature selection is to identify important features that are relevant to explain an outcome variable. Most of the work in this domain has focused on identifying globally relevant features, which are features that are related to the outcome using evidence across the entire dataset. We study a more fine-grain…
Study compares and contrasts various ML explanation methods, highlighting their disagreements and similarities.
problem Understanding and quantifying the differences between various machine learning explanation methods.
method Synthesized and visualized various explanation methods for global and local aspects of ML models.
result There is substantial agreement on the top features but less on specific rankings, and tree interpreter is comparable to SHAP for feature effects.
In many important machine learning applications, the training distribution used to learn a probabilistic classifier differs from the testing distribution on which the classifier will be used to make predictions. Traditional methods correct the distribution shift by reweighting the training data with the ratio of the de…
New method improves feature selection in tree-based models.
problem Previous feature selection methods in tree-based models lack sufficient regularization and sub-optimal performance.
method Developed a new gain penalization approach for tree-based models that allows for flexible feature-specific importance weights.
result The new method improves out-of-sample performance, especially with correlated features.
Estimating global pairwise interaction effects, i.e., the difference between the joint effect and the sum of marginal effects of two input features, with uncertainty properly quantified, is centrally important in science applications. We propose a non-parametric probabilistic method for detecting interaction effects of…