COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE AND SYSTEMS ANALYSIS

Increasing the Accuracy of Fuzzy Neural Networks by Optimizing Membership Functions

Authors

Ya. V. Ivanchuk ORCID 0000-0002-4775-6505
Vinnytsia National Technical University ROR
Yu. S. Khmelivskyi ORCID 0009-0005-3517-6748
Vinnytsia National Technical University ROR
DOI https://doi.org/10.31649/

Keywords

fuzzy neural networks membership function optimization swarm intelligence genetic algorithm time series

Abstract

The article is devoted to solving the urgent scientific and practical problem of increasing the accuracy and stability of modeling of intelligent systems based on hybrid fuzzy neural networks. The need to optimize the fuzzification parameters is due to the high sensitivity of fuzzy models to the geometric shape and localization of membership functions. The paper provides a detailed comparative analysis of the influence of membership function types (Gaussian, bell-shaped, triangular and trapezoidal), standard learning methods (hybrid algorithm, backpropagation error method) and modern population algorithms of global optimization (genetic algorithm and swarm intelligence method) on the approximation and predictive properties of a fuzzy network of the ANFIS class and a modified five-layer FNN structure. The efficiency of the studied configurations was assessed using statistical metrics of the mean square error and the coefficient of determination, as well as indicators of computational complexity and the result stability assessment. Experimental testing was performed on two different types of tasks: forecasting a complex chaotic Mackay-Glass time series and regression analysis of a real multidimensional Auto MPG dataset from the UCI Machine Learning Repository. The results obtained proved that for smooth processes without significant noise, the optimal configuration is a configuration with a Gaussian membership function and a hybrid learning method. At the same time, for real datasets, the classical hybrid method turns out to be ineffective due to falling into local minima, while the use of a genetic algorithm and swarm intelligence in combination with the Adam optimizer allows achieving stable accuracy, fast and stable convergence and significantly enhance the reliability of fuzzy neural networks functioning.

5 0

How to Cite

[1]
“Increasing the Accuracy of Fuzzy Neural Networks by Optimizing Membership Functions”, Вісник ВПІ, no. 4, pp. 141–149, Sep. 2026, doi: 10.31649/.

Author Biographies

Ya. V. Ivanchuk, Vinnytsia National Technical University

Dr. Sc. (Eng), Professor, Professor of the Chair of Computer Science

Yu. S. Khmelivskyi, Vinnytsia National Technical University

Post-Graduate Student, Assistant of the Chair of Computer Science

References

[1] Ю. С. Хмелівський та Я. В. Іванчук, «Нечіткі нейронні мережі,» на Міжнародній науково-практичній інтернет-конференції «Молодь в науці: дослідження, проблеми, перспективи (МН-2026)», Вінниця, Україна: ВНТУ, 2026. [Електронний ресурс]. Режим доступу: https://conferences.vntu.edu.ua/index.php/mn/mn2026/paper/view/27208 .
[2] A. K. S. Ahmed, M. A. Hossain, M. F. Hossain, et al., “Fuzzy neural networks and neuro-fuzzy networks: A review the main techniques and applications used in the literature,” Applied Soft Computing, vol. 95, Art. no. 106275, 2020. https://doi.org/10.1016/j.asoc.2020.106275 .
[3] Ł. Apiecionek, “Fully Scalable Fuzzy Neural Network for Data Processing,” Sensors, vol. 24, no. 16, Art. no. 5169, 2024. https://doi.org/10.3390/s24165169 .
[4] M. R. H. Adnan, A. Sarkheyli, A. M. Zain, and H. Haron, “Fuzzy logic for modeling machining process: a review,” Artificial Intelligence Review, vol. 43, no. 3, pp. 345-379, 2015. https://doi.org/10.1007/s10462-012-9381-8 .
[5] B. A. Polin, A. Rotshtein, D. Katelnikov, and O. Zelinska, “A Hybrid Fuzzy Cognitive Map and Genetic Algorithm Approach with Least-Influence Weighting for Decision-Support Forecasting,” Algorithms, vol. 19, no. 7, Art. no. 553, 2026. https://doi.org/10.3390/a19070553 .
[6] D. Karaboga and E. Kaya, “Adaptive network based fuzzy inference system (ANFIS) training approaches: a comprehensive survey,” Artificial Intelligence Review, vol. 52, no. 4, pp. 2263-2293, 2019. https://doi.org/10.1007/s10462-017-9610-2 .
[7] A. Azad, M. Manoochehri, H. Kashi, S. Farzin, H. Karami, V. Nourani, and J. Shiri, “Comparative evaluation of intelligent algorithms to improve adaptive neuro-fuzzy inference system performance in precipitation modelling,” Journal of Hydrology, vol. 571, pp. 214-224, 2019. https://doi.org/10.1016/j.jhydrol.2019.01.062 .
[8] M. Babanezhad, A. Masoumian, A. T. Nakhjiri, A. Marjani, M. Rezakazemi, and S. Shirazian, “Influence of number of membership functions on prediction of membrane systems using adaptive network based fuzzy inference system (ANFIS),” Scientific Reports, vol. 10, Art. no. 16110, 2020. https://doi.org/10.1038/s41598-020-73175-0 .
[9] K. Balasubramanian, and N. P. Ananthamoorthy, “Improved adaptive neuro-fuzzy inference system based on modified glowworm swarm and differential evolution optimization algorithm for medical diagnosis,” Neural Computing and Applications, vol. 33, no. 13, pp. 7649-7660, 2021. https://doi.org/10.1007/s00521-020-05507-0 .
[10] M. S. Hossain, S. Mekhilef, et al., “Application of adaptive neuro-fuzzy inference system for wind power density prediction,” PLOS ONE, vol. 13, no. 4, Art. no. e0193772, 2018. https://doi.org/10.1371/journal.pone.0193772 .
[11] О. В. Стародуб та В. В. Олійник, «Гібридний алгоритм навчання ANFIS-подібних нейромереж у задачах управління,» Адаптивні системи автоматичного керування, № 2 (33), с. 45-54, 2018. https://doi.org/10.20535/1560-8956.30.2017.117715 .
[12] K. O. Olatunji, S. O. Oladipo, D. M. Madyira, and Y. Sun, “Performance Evaluation of Different Clustering Techniques and Parameters of Hybrid PSO- and GA-ANFIS on Optimization and Prediction of Biomethane Yield of Alkali-Pretreated Groundnut Shells,” Waste and Biomass Valorization, vol. 16, no. 1, pp. 423-440, 2025. https://doi.org/10.1007/s12649-024-02674-2 .
[13] R. Aguilar-Rivera, M. Valenzuela-Rendón, and J. J. Rodríguez-Ortiz, “Genetic algorithms and Darwinian approaches in financial applications: a survey,” Expert Systems with Applications, vol. 42, no. 21, pp. 7684-7697, 2015. https://doi.org/10.1016/j.eswa.2015.06.001
[14] Р. Н. Квєтний, Я. В. Іванчук, І. В. Богач, О. Ю. Софина, і М. В. Барабан, Методи та алгоритми комп’ютерних обчислень. Теорія і практика. Вінниця, Україна: ВНТУ, 2023, 280 с. ISBN 978-966-641-952-4. http://ir.lib.vntu.edu.ua//handle/123456789/38075 .
[15] Я. В. Іванчук, і О. О. Борисюк, «Метаевристичний метод еволюційної оптимізації з використанням імунних підходів,» Вісник Вінницького політехнічного інституту, № 1 (184), с. 143-151, 2026. https://doi.org/10.31649/1997-9266-2026-184-1-143-151.
[16] Я. В. Іванчук, і О. О. Борисюк, «Синтез еволюційних механізмів в розробці адаптивного алгоритму оптимізації,» Проблеми програмування, № 3, с. 53-65, 2025. http://doi.org/10.15407/pp2025.03.053 .
[17] J. Yao, X. Luo, F. Li, J. Li, J. Dou, and H. Luo, “Research on hybrid strategy Particle Swarm Optimization algorithm and its applications,” Scientific Reports, vol. 14, Art. no. 24928, 2024. https://doi.org/10.1038/s41598-024-76010-y .
[18] H. Wang, et al., “AnFiS-MoH: Systematic exploration of hybrid ANFIS frameworks via metaheuristic optimization hybridization with evolutionary and swarm-based algorithms,” Applied Soft Computing, vol. 167, Part A, Art. no. 112334, 2024. https://doi.org/10.1016/j.asoc.2024.112334 .
[19] T. J. Ross, Fuzzy Logic with Engineering Applications, 4th ed. Hoboken, NJ, USA: Wiley, 2016. ISBN 978-1-119-23586-0. https://home.iitk.ac.in/~avrs/ManyValuedLogic/FuzzyLogicforEngineers.pdf .
[20] UCI Machine Learning Repository, “Auto MPG Data Set,” University of California, Irvine, 2023. [Online]. Available: https://archive.ics.uci.edu/dataset/9/auto+mpg .

Most read articles by the same author(s)