Factual and counterfactual explanations in fuzzy classification trees

Classification algorithms have recently acquired great popularity due to their efficiency to generate models capable of solving high complexity problems. Specifically, black box models are the ones that offer the best results, since they greatly benefit from the enormous amount of data available to...

Descripción completa

Detalles Bibliográficos
Autores: Fernández Martín, Guillermo Tomás, Aledo Sánchez, Juan Ángel, Gámez Martín, José Antonio, Puerta Callejón, José Miguel
Tipo de recurso: artículo
Fecha de publicación:2022
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/44493
Acceso en línea:http://dx.doi.org/10.1109/tfuzz.2022.3179582
https://hdl.handle.net/10578/44493
Access Level:acceso abierto
Palabra clave:Counterfactual explanations
Explainable artificial intelligence (XAI)
Factual explanations
Fuzzy decision trees
Fuzzy reasoning
Robustness
id ES_2d18b4af25abdffa92ffcfda311eb5a1
oai_identifier_str oai:ruidera.uclm.es:10578/44493
network_acronym_str ES
network_name_str España
repository_id_str
spelling Factual and counterfactual explanations in fuzzy classification treesFernández Martín, Guillermo TomásAledo Sánchez, Juan ÁngelGámez Martín, José AntonioPuerta Callejón, José MiguelCounterfactual explanationsExplainable artificial intelligence (XAI)Factual explanationsFuzzy decision treesFuzzy reasoningRobustnessClassification algorithms have recently acquired great popularity due to their efficiency to generate models capable of solving high complexity problems. Specifically, black box models are the ones that offer the best results, since they greatly benefit from the enormous amount of data available to learn models in an increasingly accurate way. However, their main disadvantage compared to other simpler algorithms, e.g., decision trees, is the loss of interpretability for both the model and the individual classifications, which may become a major drawback because of the increasing number of applications where it is advisable and even compulsory to provide an explanation. A well-accepted practice is to build an explainable model that can mimic the behavior of the (more complex) classifier in the neighborhood of the instance to be explained. Nonetheless, the generation of explanations in such white box models is not trivial either, which has generated intense research. It is common to generate two types of explanations, factual explanations and counterfactual explanations, which complement each other to justify why the instance has been classified into a certain class or category. In this work, we propose the definition of factual and counterfactual explanations in the frame of fuzzy decision trees, where multiple branches can be fired at once. Our proposal is centered around the definition of factual explanations that can contain more than a single rule, in contrast to the current standard that is limited to considering a single rule as a factual explanation. Moreover, we introduce the idea of robust factual explanation. Finally, we provide procedures to obtain counterfactual explanations from the instance and also from a factual explanation.IEEE Press202520252022info:eu-repo/semantics/articleapplication/pdfapplication/pdfhttp://dx.doi.org/10.1109/tfuzz.2022.3179582https://hdl.handle.net/10578/44493reponame:RUIdeRA. Repositorio Institucional de la UCLMinstname:Consejo Superior de Investigaciones Científicas (CSIC)InglésPID2019–106758GB–C33AEI/ 10.13039/501100011033SBPLY/17/180501/000493FPU19/02930info:eu-repo/semantics/openAccessoai:ruidera.uclm.es:10578/444932026-05-27T07:36:41Z
dc.title.none.fl_str_mv Factual and counterfactual explanations in fuzzy classification trees
title Factual and counterfactual explanations in fuzzy classification trees
spellingShingle Factual and counterfactual explanations in fuzzy classification trees
Fernández Martín, Guillermo Tomás
Counterfactual explanations
Explainable artificial intelligence (XAI)
Factual explanations
Fuzzy decision trees
Fuzzy reasoning
Robustness
title_short Factual and counterfactual explanations in fuzzy classification trees
title_full Factual and counterfactual explanations in fuzzy classification trees
title_fullStr Factual and counterfactual explanations in fuzzy classification trees
title_full_unstemmed Factual and counterfactual explanations in fuzzy classification trees
title_sort Factual and counterfactual explanations in fuzzy classification trees
dc.creator.none.fl_str_mv Fernández Martín, Guillermo Tomás
Aledo Sánchez, Juan Ángel
Gámez Martín, José Antonio
Puerta Callejón, José Miguel
author Fernández Martín, Guillermo Tomás
author_facet Fernández Martín, Guillermo Tomás
Aledo Sánchez, Juan Ángel
Gámez Martín, José Antonio
Puerta Callejón, José Miguel
author_role author
author2 Aledo Sánchez, Juan Ángel
Gámez Martín, José Antonio
Puerta Callejón, José Miguel
author2_role author
author
author
dc.subject.none.fl_str_mv Counterfactual explanations
Explainable artificial intelligence (XAI)
Factual explanations
Fuzzy decision trees
Fuzzy reasoning
Robustness
topic Counterfactual explanations
Explainable artificial intelligence (XAI)
Factual explanations
Fuzzy decision trees
Fuzzy reasoning
Robustness
description Classification algorithms have recently acquired great popularity due to their efficiency to generate models capable of solving high complexity problems. Specifically, black box models are the ones that offer the best results, since they greatly benefit from the enormous amount of data available to learn models in an increasingly accurate way. However, their main disadvantage compared to other simpler algorithms, e.g., decision trees, is the loss of interpretability for both the model and the individual classifications, which may become a major drawback because of the increasing number of applications where it is advisable and even compulsory to provide an explanation. A well-accepted practice is to build an explainable model that can mimic the behavior of the (more complex) classifier in the neighborhood of the instance to be explained. Nonetheless, the generation of explanations in such white box models is not trivial either, which has generated intense research. It is common to generate two types of explanations, factual explanations and counterfactual explanations, which complement each other to justify why the instance has been classified into a certain class or category. In this work, we propose the definition of factual and counterfactual explanations in the frame of fuzzy decision trees, where multiple branches can be fired at once. Our proposal is centered around the definition of factual explanations that can contain more than a single rule, in contrast to the current standard that is limited to considering a single rule as a factual explanation. Moreover, we introduce the idea of robust factual explanation. Finally, we provide procedures to obtain counterfactual explanations from the instance and also from a factual explanation.
publishDate 2022
dc.date.none.fl_str_mv 2022
2025
2025
dc.type.none.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv http://dx.doi.org/10.1109/tfuzz.2022.3179582
https://hdl.handle.net/10578/44493
url http://dx.doi.org/10.1109/tfuzz.2022.3179582
https://hdl.handle.net/10578/44493
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv PID2019–106758GB–C33
AEI/ 10.13039/501100011033
SBPLY/17/180501/000493
FPU19/02930
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv IEEE Press
publisher.none.fl_str_mv IEEE Press
dc.source.none.fl_str_mv reponame:RUIdeRA. Repositorio Institucional de la UCLM
instname:Consejo Superior de Investigaciones Científicas (CSIC)
instname_str Consejo Superior de Investigaciones Científicas (CSIC)
reponame_str RUIdeRA. Repositorio Institucional de la UCLM
collection RUIdeRA. Repositorio Institucional de la UCLM
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869405287033602048
score 15,81155