EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

FuzzyGNN: a graph neural network framework for enhanced fuzzy modeling

Gianluca Zaza, Raffaele Scaringi, Gennaro Vessio, Giovanna Castellano

Neural Computing and Applications · 2026-05 · 38 (10)

DOI: 10.1007/s00521-026-11967-7 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neural Computing and Applications, online erschienen am 09.05.2026. Verfasst von Gianluca Zaza und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Università degli Studi di Bari Aldo Moro.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Balancing predictive performance with transparency remains a core challenge in eXplainable Artificial Intelligence, especially for tabular data. In this work, we present FuzzyGNN, a novel hybrid framework that combines fuzzy logic with Graph Neural Networks (GNNs) to deliver both high accuracy and interpretability. FuzzyGNN represents each data instance as a graph, where raw features and fuzzy sets are modeled as nodes, enabling contextual reasoning through message passing. Unlike traditional neuro-fuzzy systems—often limited by scalability and rule explosion—FuzzyGNN enforces model compactness through its graph-aware architecture and a pruning mechanism guided by graphbased explanations. Extensive experiments on seven benchmark datasets show that FuzzyGNN achieves competitive predictive performance while drastically reducing the number of fuzzy rules, linguistic terms, and active features used to produce output fuzzy rules. Additionally, it offers structural interpretability through rule extraction and post-hoc explainability via GNNExplainer. These capabilities make FuzzyGNN particularly well-suited for domains where transparency is as crucial as accuracy, such as healthcare, finance, and safety-critical systems. The code is available in the following link: https://github.com/cilabuniba/fuzzygnn

Förderung

  • Università degli Studi di Bari Aldo Moro

Zitieren

APA

Zaza, G., Scaringi, R., Vessio, G., & Castellano, G. (2026). FuzzyGNN: a graph neural network framework for enhanced fuzzy modeling. Neural Computing and Applications, 38(10). https://doi.org/10.1007/s00521-026-11967-7

Vancouver

Zaza G, Scaringi R, Vessio G, Castellano G. FuzzyGNN: a graph neural network framework for enhanced fuzzy modeling. Neural Computing and Applications. 2026;38(10). doi:10.1007/s00521-026-11967-7

BibTeX
@article{Zaza2026,
  author = {Zaza, Gianluca and Scaringi, Raffaele and Vessio, Gennaro and Castellano, Giovanna},
  title = {FuzzyGNN: a graph neural network framework for enhanced fuzzy modeling},
  journal = {Neural Computing and Applications},
  year = {2026},
  volume = {38},
  number = {10},
  doi = {10.1007/s00521-026-11967-7}
}
Kostenlos starten →