The Truth is in There : Rule Extraction from Opaque Models Using Genetic Programming

By Ulf Johansson, Rikard König, Lars Niklasson. 2004

65 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability, Evolutionary Computation

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

In Evolutionary Computation

Citations in the corpus

Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI (2019, positive): “Although it was not originally intended for extracting rules from opaque models, the generic proposition of G-REX has been extended to also account for model explainability purposes [190,211] .”
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) (2018, background): “One effort to gain insight into highly complex models is the use of rule extraction [114]–[116].”
A Survey of Methods for Explaining Black Box Models (2018, methodology): “Reference [45] extends G-REX for handling regression problems by generating regression trees, and classification problems by generating fuzzy rules.”
Comprehensible credit scoring models using rule extraction from support vector machines (2007, neutral): “G-REX A technique recently suggested, named G-REX (Genetic Rule EXtraction) [16], is a pedagogical method to extract rules from artificial neural networks with the use of genetic programming [18], which is based on Darwin’s principle of ‘survival of the fittest’.”
Interpreting Black-Box Models: A Review on Explainable Artificial Intelligence (2023, methodology): “To handle regression and classification issues using regression trees and fuzzy rules, respectively, the authors further created G-REX [99].”
Explainable Artificial Intelligence: a Systematic Review (2020, methodology): “Similarly, Genetic Rule EXtraction (G-REX) [166, 167] employed genetic algorithms to generate IF-THEN rules with AND/OR operators.”
A Diagnostic Study of Explainability Techniques for Text Classification (2020, background): “They can further be categorised as: employing model gradients (Sundararajan et al., 2017; Simonyan et al., 2013), being perturbation based (Shapley, 1953; Zeiler and Fergus, 2014) or providing explanations through model simpliﬁcations (Ribeiro et al., 2016; Johansson et al., 2004).”
Extended Taxonomy of Rule Extraction Techniques and Assessment of KDRuleEx (2012, background): “To answer these problems one may start from a less accurate simple description of the data and then can transform it into more accurate complex description rules or may start with accurate opaque model and can transform into more comprehensible description [36] [37].”
KDRuleEx: A Novel Approach for Enhancing User Comprehensibility Using Rule Extraction (2012, methodology): “Transparent Model derived using dataset and opaque model may have better accuracy then model derived directly data [3] [4] [5].”
A Survey on the Explainability of Supervised Machine Learning (2020, methodology): “REx Genetic algorithm Classification Pedagogical Rule-based Kamruzzaman (2010) GEX Genetic algorithm Classification Pedagogical Rule-based Markowska-Kaczmar and Chumieja (2004) DeepRED Tree induction Classification Decompositional [ANN] Rule-based Zilke et al. (2016) SUBSET Feature Importance Classi”
Interpretable Rule Discovery Through Bilevel Optimization of Split-Rules of Nonlinear Decision Trees for Classification Problems (2020, methodology): “Authors in [16] use a pedagogical technique to evolve comprehensible rules using genetic-programming.”
Rule Extraction from Support Vector Machines: An Overview of Issues and Application in Credit Scoring (2008, methodology): “An interesting observation is that the (generally) better performing non-linear model can be used in a pre-processing step to clean up the data [35,42].”