Rule Extraction in Trained Feedforward Deep Neural Networks: Integrating Cosine Similarity and Logic for Explainability

cic.institucionOrigenLaboratorio de Investigación y Formación en Informática Avanzada (LIFIA)
cic.isFulltextSI
cic.isPeerReviewedSI
cic.lugarDesarrolloLaboratorio de Investigación y Formación en Informática Avanzada (LIFIA)
cic.parentTypeArtículo
cic.versionPublicada
dc.date.accessioned2025-04-24T12:48:00Z
dc.date.available2025-04-24T12:48:00Z
dc.identifier.urihttps://digital.cic.gba.gob.ar/handle/11746/12471
dc.titleRule Extraction in Trained Feedforward Deep Neural Networks: Integrating Cosine Similarity and Logic for Explainabilityen
dc.typeArtículo
dcterms.abstractExplainability is a key aspect of machine learning, necessary for ensuring transparency and trust in decision-making processes. As machine learning models become more complex, the integration of neural and symbolic approaches has emerged as a promising solution to the explainability problem. One effective solution involves using search techniques to extract rules from trained deep neural networks by examining weight and bias values and calculating their correlation with outputs. This article proposes incorporating cosine similarity in this process to narrow down the search space and identify the critical path connecting inputs to final results. Additionally, the integration of first-order logic (FOL) is suggested to provide a more comprehensive and interpretable understanding of the decision-making process. By leveraging cosine similarity and FOL, an innovative algorithm capable of extracting and explaining rule patterns learned by a feedforward trained neural network was developed and tested in two use cases, demonstrating its effectiveness in providing insights into model behavior.en
dcterms.creator.authorNegro, Pablo Ariel
dcterms.creator.authorPons, Claudia Fabiana
dcterms.identifier.otherISSN: 2642-1585
dcterms.identifier.otherDOI: 10.4018/IJAIML.347988
dcterms.identifier.urlhttps://doi.org/10.4018/IJAIML.347988
dcterms.isPartOf.issuevol. 13, no. 1
dcterms.isPartOf.seriesInternational Journal of Artificial Intelligence and Machine Learning
dcterms.issued2024-12
dcterms.languageInglés
dcterms.licenseAttribution 4.0 International (BY 4.0)
dcterms.subjectArtificial Intelligenceen
dcterms.subjectBlack Box Modelsen
dcterms.subjectCosine Similarityen
dcterms.subjectDeep Learningen
dcterms.subjectDistance Functionen
dcterms.subjectEntropyen
dcterms.subjectExplainabilityen
dcterms.subjectFeedforward Neural Networken
dcterms.subjectLogicen
dcterms.subjectRegularizationen
dcterms.subjectRule Extractionen
dcterms.subject.materiaCiencias de la Computación e Información

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
Rule Extraction in Trained.pdf-PDFA.pdf
Tamaño:
1019.65 KB
Formato:
Adobe Portable Document Format
Descripción:
Documento completo

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
3.46 KB
Formato:
Item-specific license agreed upon to submission
Descripción: