77706

Автор(ы): 

Автор(ов): 

5

Параметры публикации

Тип публикации: 

Доклад

Название: 

Neuro-Fuzzy Model Based on Multidimensional Membership Function

ISBN/ISSN: 

978-3-031-30647-1 ISBN 978-3-031-30648-8 (eBook)

DOI: 

10.1007/978-3-031-30648-8

Наименование конференции: 

  • 25th International Conference on Distributed Computer and Communication Networks: Control, Computation, Communications (DCCN-2022)

Наименование источника: 

  • Lecture Note Computer Sciences (25th International Conference DCCN2022 Moscow)

Город: 

  • Moscow

Издательство: 

  • Springer

Год издания: 

2022

Страницы: 

234-245
Аннотация
Currently, systems based on multidimensional membership functions are being actively researched and developed. Most algorithms for determining the parameters of fuzzy membership functions are developed on the basis of one-dimensional membership functions. The fuzzy rules generated by these algorithms often overlap and cannot act as independent rules. The overlap of fuzzy rules in fuzzy systems does not allow one to evaluate the reliability of individual fuzzy rules and at the same time creates limitations in extracting knowledge from fuzzy systems. In this article, a neuro-fuzzy neural system will be built based on a multidimensional Gaussian membership function with the ability to describe the relationship of interaction between input variables, and at the same time, the generated fuzzy rules are capable of independent operation.

Библиографическая ссылка: 

Буй З.Т., Пащенко Ф.Ф., Хиеу Н.В., Нгуен В.Ч., Нгуен Ф.Т. Neuro-Fuzzy Model Based on Multidimensional Membership Function / Lecture Note Computer Sciences (25th International Conference DCCN2022 Moscow). Moscow: Springer, 2022. С. 234-245.