EUVIMEDEUVIMED entdecken
Crossref

Fuzzy Multiple Linear Least Squares Regression Analysis in the Hilbert Space of A-Linearly Correlated Fuzzy Numbers

Laecio Carvalho Barros, Kadu Vinicius Toledo Paulino, Estevao Esmi, Francielle Santo Pedro Simoes, Otavio Akira de Barros, Vinicius Francisco Wasques

IEEE Transactions on Fuzzy Systems · 2026-08 · 34 (8) · 2641-2650

DOI: 10.1109/tfuzz.2026.3697972 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Fuzzy Systems, 2026-08. Verfasst von Laecio Carvalho Barros und 5 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Gefördert von: Conselho Nacional de Desenvolvimento Científico e Tecnológico, Fundação de Amparo à Pesquisa do Estado de São Paulo, Coordenação de Aperfeiçoamento de Pessoal de Nível Superior.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This article presents a fuzzy least squares method for multiple linear regression models formulated in the space of $A$ -linearly correlated fuzzy numbers, denoted by $\mathbb {R}_{\mathcal {F}(A)}$ . The algebraic operations in $\mathbb {R}_{\mathcal {F}(A)}$ are defined through an isometric isomorphism with the complex field $\mathbb {C}$ , guaranteeing that $\mathbb {R}_{\mathcal {F}(A)}$ is a field and that $\mathbb {R}_{\mathcal {F}(A)}^{p}$ is a Hilbert space over $\mathbb {R}_{\mathcal {F}(A)}$ . As a consequence, all the tools of differential and integral calculus for functions with fuzzy input and output become available, as well as the full framework of orthogonal projection theory in this class of fuzzy numbers. Based on these fundamentals, we propose a least squares method for fuzzy multiple linear regression models, where the coefficients can be either real or fuzzy numbers. Numerical results show that the proposed $\mathbb {R}_{\mathcal {F}(A)}$ -based formulation yields lower root-mean-squared error values in comparison with existing approaches, lower mean absolute error values under the $\mathbb {R}_{\mathcal {F}(A)}$ norm, and higher agreement with the observed fuzzy outputs, as measured by similarity.

Förderung

  • Conselho Nacional de Desenvolvimento Científico e Tecnológico
  • Fundação de Amparo à Pesquisa do Estado de São Paulo
  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Zitieren

APA

Barros, L. C., Paulino, K. V. T., Esmi, E., Simoes, F. S. P., Barros, O. A. D., & Wasques, V. F. (2026). Fuzzy Multiple Linear Least Squares Regression Analysis in the Hilbert Space of A-Linearly Correlated Fuzzy Numbers. IEEE Transactions on Fuzzy Systems, 34(8), 2641-2650. https://doi.org/10.1109/tfuzz.2026.3697972

Vancouver

Barros LC, Paulino KVT, Esmi E, Simoes FSP, Barros OAD, Wasques VF. Fuzzy Multiple Linear Least Squares Regression Analysis in the Hilbert Space of A-Linearly Correlated Fuzzy Numbers. IEEE Transactions on Fuzzy Systems. 2026;34(8):2641-2650. doi:10.1109/tfuzz.2026.3697972

BibTeX
@article{Barros2026,
  author = {Barros, Laecio Carvalho and Paulino, Kadu Vinicius Toledo and Esmi, Estevao and Simoes, Francielle Santo Pedro and Barros, Otavio Akira de and Wasques, Vinicius Francisco},
  title = {Fuzzy Multiple Linear Least Squares Regression Analysis in the Hilbert Space of A-Linearly Correlated Fuzzy Numbers},
  journal = {IEEE Transactions on Fuzzy Systems},
  year = {2026},
  volume = {34},
  number = {8},
  pages = {2641--2650},
  doi = {10.1109/tfuzz.2026.3697972}
}
Kostenlos starten →