TY - GEN
T1 - EvoNFuzz: A New Evolutionary Neuro-Fuzzy Network with Genetic Programming-Based Learning
AU - Brás, Glender
AU - Silva, Alisson Marques
AU - Wanner, Elizabeth F.
N1 - Copyright © 2026 The Author(s), under exclusive license to Springer Nature Switzerland AG. This version of the article has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use [ https://www.springernature.com/gp/open-research/policies/accepted-manuscript-terms ] but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/978-3-032-05179-0_37
PY - 2026
Y1 - 2026
N2 - This paper presents EvoNFuzz, a novel Evolutionary Neuro-Fuzzy Network that integrates functional fuzzy rules with a hybrid learning approach combining Multi-Gene Genetic Programming (MGGP) and gradient-based optimization. Unlike traditional Takagi-Sugeno models, EvoNFuzz employs polynomial-based consequents, evolved via MGGP, to more effectively capture complex non-linear relationships in data. Additionally, EvoNFuzz incorporates rule weights akin to those in neural networks, allowing it to assign varying degrees of importance to each fuzzy rule. The membership functions are determined using the K-Means clustering algorithm. A Gradient-based learning algorithm adjusts the rule weights and the membership functions. The performance of EvoNFuzz is rigorously tested against alternative models on non-linear regression tasks. The computational results demonstrate that EvoNFuzz consistently outperforms or matches the performance of alternative models.
AB - This paper presents EvoNFuzz, a novel Evolutionary Neuro-Fuzzy Network that integrates functional fuzzy rules with a hybrid learning approach combining Multi-Gene Genetic Programming (MGGP) and gradient-based optimization. Unlike traditional Takagi-Sugeno models, EvoNFuzz employs polynomial-based consequents, evolved via MGGP, to more effectively capture complex non-linear relationships in data. Additionally, EvoNFuzz incorporates rule weights akin to those in neural networks, allowing it to assign varying degrees of importance to each fuzzy rule. The membership functions are determined using the K-Means clustering algorithm. A Gradient-based learning algorithm adjusts the rule weights and the membership functions. The performance of EvoNFuzz is rigorously tested against alternative models on non-linear regression tasks. The computational results demonstrate that EvoNFuzz consistently outperforms or matches the performance of alternative models.
UR - https://link.springer.com/chapter/10.1007/978-3-032-05179-0_37
UR - https://www.scopus.com/pages/publications/105024539298
U2 - 10.1007/978-3-032-05179-0_37
DO - 10.1007/978-3-032-05179-0_37
M3 - Conference publication
SN - 9783032051783
T3 - Lecture Notes in Computer Science (LNCS)
SP - 493
EP - 505
BT - Progress in Artificial Intelligence
A2 - Valente de Oliveira, José
A2 - Rodrigues, João
A2 - Dias, João
A2 - Cardoso, Pedro
A2 - Leite, João
PB - Springer, Cham
ER -