A New Method to Compare the Interpretability of Rule-Based Algorithms
Interpretability is becoming increasingly important for predictive model analysis. Unfortunately, as remarked by many authors, there is still no consensus regarding this notion. The goal of this paper is to propose the definition of a score that allows for quickly comparing interpretable algorithms....
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-11-01
|
Series: | AI |
Subjects: | |
Online Access: | https://www.mdpi.com/2673-2688/2/4/37 |
_version_ | 1797506987542446080 |
---|---|
author | Vincent Margot George Luta |
author_facet | Vincent Margot George Luta |
author_sort | Vincent Margot |
collection | DOAJ |
description | Interpretability is becoming increasingly important for predictive model analysis. Unfortunately, as remarked by many authors, there is still no consensus regarding this notion. The goal of this paper is to propose the definition of a score that allows for quickly comparing interpretable algorithms. This definition consists of three terms, each one being quantitatively measured with a simple formula: <i>predictivity</i>, <i>stability</i> and <i>simplicity</i>. While predictivity has been extensively studied to measure the accuracy of predictive algorithms, stability is based on the Dice-Sorensen index for comparing two rule sets generated by an algorithm using two independent samples. The simplicity is based on the sum of the lengths of the rules derived from the predictive model. The proposed score is a weighted sum of the three terms mentioned above. We use this score to compare the interpretability of a set of rule-based algorithms and tree-based algorithms for the regression case and for the classification case. |
first_indexed | 2024-03-10T04:40:17Z |
format | Article |
id | doaj.art-2e9f39c6ea234a87b0cb6f6583f889ed |
institution | Directory Open Access Journal |
issn | 2673-2688 |
language | English |
last_indexed | 2024-03-10T04:40:17Z |
publishDate | 2021-11-01 |
publisher | MDPI AG |
record_format | Article |
series | AI |
spelling | doaj.art-2e9f39c6ea234a87b0cb6f6583f889ed2023-11-23T03:24:32ZengMDPI AGAI2673-26882021-11-012462163510.3390/ai2040037A New Method to Compare the Interpretability of Rule-Based AlgorithmsVincent Margot0George Luta1Advestis, 69 Boulevard Haussmann, F-75008 Paris, FranceDepartment of Biostatistics, Bioinformatics and Biomathematics, Georgetown University, Washington, DC 20057, USAInterpretability is becoming increasingly important for predictive model analysis. Unfortunately, as remarked by many authors, there is still no consensus regarding this notion. The goal of this paper is to propose the definition of a score that allows for quickly comparing interpretable algorithms. This definition consists of three terms, each one being quantitatively measured with a simple formula: <i>predictivity</i>, <i>stability</i> and <i>simplicity</i>. While predictivity has been extensively studied to measure the accuracy of predictive algorithms, stability is based on the Dice-Sorensen index for comparing two rule sets generated by an algorithm using two independent samples. The simplicity is based on the sum of the lengths of the rules derived from the predictive model. The proposed score is a weighted sum of the three terms mentioned above. We use this score to compare the interpretability of a set of rule-based algorithms and tree-based algorithms for the regression case and for the classification case.https://www.mdpi.com/2673-2688/2/4/37interpretabilitytransparencyexplainabilitypredictivitystabilitysimplicity |
spellingShingle | Vincent Margot George Luta A New Method to Compare the Interpretability of Rule-Based Algorithms AI interpretability transparency explainability predictivity stability simplicity |
title | A New Method to Compare the Interpretability of Rule-Based Algorithms |
title_full | A New Method to Compare the Interpretability of Rule-Based Algorithms |
title_fullStr | A New Method to Compare the Interpretability of Rule-Based Algorithms |
title_full_unstemmed | A New Method to Compare the Interpretability of Rule-Based Algorithms |
title_short | A New Method to Compare the Interpretability of Rule-Based Algorithms |
title_sort | new method to compare the interpretability of rule based algorithms |
topic | interpretability transparency explainability predictivity stability simplicity |
url | https://www.mdpi.com/2673-2688/2/4/37 |
work_keys_str_mv | AT vincentmargot anewmethodtocomparetheinterpretabilityofrulebasedalgorithms AT georgeluta anewmethodtocomparetheinterpretabilityofrulebasedalgorithms AT vincentmargot newmethodtocomparetheinterpretabilityofrulebasedalgorithms AT georgeluta newmethodtocomparetheinterpretabilityofrulebasedalgorithms |