A narrative review on the interpretability of fuzzy rule-based models from a modern interpretable machine learning perspective
- Author
- Lynn Pickering (UGent) , Kelly Cohen and Bernard De Baets (UGent)
- Organization
- Project
- Abstract
- Interpretable machine learning is an increasingly popular field as machine learning grows in prevalence and power. Though the fuzzy (rule-based) modeling field has long considered interpretability to be a great strength of its models, fuzzy rule-based models are often missing in the mainstream discussion about interpretable machine learning. The original contribution of this work is to uncover the reasons for this absence, and to propose actionable steps so that the fuzzy modeling field may join this mainstream discussion. To do so, we perform a narrative review by analyzing a select number of the most highly cited and broad papers in the interpretable machine learning and fuzzy modeling fields through three lenses of interpretability: scale, audience, and stage dependence. We identify several gaps in the fuzzy modeling interpretability discussion and give recommendations for future work in each lens. Based on the lenses of interpretability, we provide a simple manner in which to analyze the interpretability of a model that should be used in future works when discussing the interpretability of a model. Building interpretable models that humans can understand is an important task, and this analysis will help bring another beneficial voice to the mainstream conversation.
- Keywords
- Interpretability, Fuzzy rule-based models, Machine learning, Explainable artificial, intelligence, INFERENCE SYSTEMS, BLACK-BOX
Downloads
-
s40815-025-02022-z.pdf
- full text (Published version)
- |
- open access
- |
- |
- 2.40 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01KVSV580E03KWXM89DZWPHNTS
- MLA
- Pickering, Lynn, et al. “A Narrative Review on the Interpretability of Fuzzy Rule-Based Models from a Modern Interpretable Machine Learning Perspective.” INTERNATIONAL JOURNAL OF FUZZY SYSTEMS, vol. 28, no. 4, 2025, pp. 1313–32, doi:10.1007/s40815-025-02022-z.
- APA
- Pickering, L., Cohen, K., & De Baets, B. (2025). A narrative review on the interpretability of fuzzy rule-based models from a modern interpretable machine learning perspective. INTERNATIONAL JOURNAL OF FUZZY SYSTEMS, 28(4), 1313–1332. https://doi.org/10.1007/s40815-025-02022-z
- Chicago author-date
- Pickering, Lynn, Kelly Cohen, and Bernard De Baets. 2025. “A Narrative Review on the Interpretability of Fuzzy Rule-Based Models from a Modern Interpretable Machine Learning Perspective.” INTERNATIONAL JOURNAL OF FUZZY SYSTEMS 28 (4): 1313–32. https://doi.org/10.1007/s40815-025-02022-z.
- Chicago author-date (all authors)
- Pickering, Lynn, Kelly Cohen, and Bernard De Baets. 2025. “A Narrative Review on the Interpretability of Fuzzy Rule-Based Models from a Modern Interpretable Machine Learning Perspective.” INTERNATIONAL JOURNAL OF FUZZY SYSTEMS 28 (4): 1313–1332. doi:10.1007/s40815-025-02022-z.
- Vancouver
- 1.Pickering L, Cohen K, De Baets B. A narrative review on the interpretability of fuzzy rule-based models from a modern interpretable machine learning perspective. INTERNATIONAL JOURNAL OF FUZZY SYSTEMS. 2025;28(4):1313–32.
- IEEE
- [1]L. Pickering, K. Cohen, and B. De Baets, “A narrative review on the interpretability of fuzzy rule-based models from a modern interpretable machine learning perspective,” INTERNATIONAL JOURNAL OF FUZZY SYSTEMS, vol. 28, no. 4, pp. 1313–1332, 2025.
@article{01KVSV580E03KWXM89DZWPHNTS,
abstract = {{Interpretable machine learning is an increasingly popular field as machine learning grows in prevalence and power. Though the fuzzy (rule-based) modeling field has long considered interpretability to be a great strength of its models, fuzzy rule-based models are often missing in the mainstream discussion about interpretable machine learning. The original contribution of this work is to uncover the reasons for this absence, and to propose actionable steps so that the fuzzy modeling field may join this mainstream discussion. To do so, we perform a narrative review by analyzing a select number of the most highly cited and broad papers in the interpretable machine learning and fuzzy modeling fields through three lenses of interpretability: scale, audience, and stage dependence. We identify several gaps in the fuzzy modeling interpretability discussion and give recommendations for future work in each lens. Based on the lenses of interpretability, we provide a simple manner in which to analyze the interpretability of a model that should be used in future works when discussing the interpretability of a model. Building interpretable models that humans can understand is an important task, and this analysis will help bring another beneficial voice to the mainstream conversation.}},
author = {{Pickering, Lynn and Cohen, Kelly and De Baets, Bernard}},
issn = {{1562-2479}},
journal = {{INTERNATIONAL JOURNAL OF FUZZY SYSTEMS}},
keywords = {{Interpretability,Fuzzy rule-based models,Machine learning,Explainable artificial,intelligence,INFERENCE SYSTEMS,BLACK-BOX}},
language = {{eng}},
number = {{4}},
pages = {{1313--1332}},
title = {{A narrative review on the interpretability of fuzzy rule-based models from a modern interpretable machine learning perspective}},
url = {{http://doi.org/10.1007/s40815-025-02022-z}},
volume = {{28}},
year = {{2025}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: