Model Class Reliance: Variable Importance Measures for any Machine Learning Model Class, from the "Rashomon" Perspective

By Aaron Fisher, Cynthia Rudin, Francesca Dominici. 2018

187 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

FairSight: Visual Analytics for Fairness in Decision Making (2019, background): “For example, feature-level auditing methods seek to analyze the feature importance by permutation [11] or quantifying feature interaction [6, 13, 25].”
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) (2018, background): “[113] and [132]).”
The age of secrecy and unfairness in recidivism prediction (2018, background): “The result of (29) makes no such assumption, and does not compute significance.”
A Survey on the Explainability of Supervised Machine Learning (2020, methodology): “Freitas (2014) suggests a different measure for the complexity of a rule-based model.”
Are Explanations Helpful? A Comparative Study of the Effects of Explanations in AI-Assisted Decision-Making (2021, methodology): “Specifically, we adopted the permutation feature importance method [28] to compute the importance of each feature as the increase of the model’s prediction error after permuting the values on that feature, and we visualized different feature’s importance using a bar chart (Figure 1(b)).”
Fairness and Transparency of Machine Learning for Trustworthy Cloud Services (2018, background): “…of this kind of approach are: (i) feature importance, that measures in a model agnostic way the importance of features (which are used for explanations) [21], and (ii) local surrogate models, that ﬁt local, interpretable models that can explain single predictions of any machine learning model [22].”
Variable Importance Clouds: A Way to Explore Variable Importance for the Set of Good Models (2019, methodology): “Model reliance function MR in Definition 2.3 turns out to have a specific formula for ridge regression models, given by the lemma below, which is a generalization of Theorem 2 of Fisher et al. (2018) to ridge regression.”
Machine Learning Interpretability: A Survey on Methods and Metrics (2019, background): “Partial Dependence Plot [142] Global Feature summary Individual Condition Expectation [143] Global/Local Feature summary Accumulated Local Effects Plot [144] Global Feature summary Feature Interaction [145] Global Feature summary Feature Importance [146] Global/Local Feature summary Local Surrogate ”
Explainable AI (XAI): Core Ideas, Techniques, and Solutions (2022, background): “Feature importance [17, 19] refers to a class of techniques that are used to assign scores to input features.”
Machine Learning-Based Network Vulnerability Analysis of Industrial Internet of Things (2019, background): “For more detailed information, we refer readers to [38] and [39].”