Simple heuristic approach for training of type-2 NEO-fuzzy neural network

Yancho Todorov, Margarita Terziyska

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

Abstract

This paper describes the development of Interval Type-2 NEO-Fuzzy Neural Network for modeling of complex dynamics. The proposed network represents a parallel set of multiple zero order Sugeno type approximations, related only to their own input argument. As learning procedure a simple heuristic backpropagation approach, where the sign of the gradient is taken into account, is adopted. To improve the robustness of the network and the possibilities for handling uncertainties, Interval Type-2 Gaussian fuzzy sets are introduced into the network topology. The potentials of the proposed approach in modeling of Mackey-Glass and Rossler Chaotic time series are studied. A comparison is made with the classical Gradient Descent learning approach.

Original languageEnglish
Title of host publicationProceedings of the 2015 20th International Conference on Process Control, PC 2015
EditorsM. Fikar, M. Kvasnica
PublisherInstitute of Electrical and Electronic Engineers IEEE
Pages278-283
ISBN (Print)978-1-4673-6627-4
DOIs
Publication statusPublished - 28 Jul 2015
MoE publication typeA4 Article in a conference publication
Event20th International Conference on Process Control, PC 2015 - Strbske Pleso, Slovakia
Duration: 9 Jun 201512 Jun 2015

Conference

Conference20th International Conference on Process Control, PC 2015
CountrySlovakia
CityStrbske Pleso
Period9/06/1512/06/15

Fingerprint

Fuzzy neural networks
Fuzzy Neural Network
Fuzzy sets
Backpropagation
Time series
Topology
Heuristics
Multiple Zeros
Glass
Chaotic Time Series
Interval
Gradient Descent
Back Propagation
Complex Dynamics
Modeling
Network Topology
Fuzzy Sets
Gradient
Robustness
Uncertainty

Keywords

  • Chaotic time-series prediction
  • Dynamic modeling
  • Fuzzy systems
  • Neo-fuzzy neuron
  • Neural networks

Cite this

Todorov, Y., & Terziyska, M. (2015). Simple heuristic approach for training of type-2 NEO-fuzzy neural network. In M. Fikar, & M. Kvasnica (Eds.), Proceedings of the 2015 20th International Conference on Process Control, PC 2015 (pp. 278-283). [7169976] Institute of Electrical and Electronic Engineers IEEE. https://doi.org/10.1109/PC.2015.7169976
Todorov, Yancho ; Terziyska, Margarita. / Simple heuristic approach for training of type-2 NEO-fuzzy neural network. Proceedings of the 2015 20th International Conference on Process Control, PC 2015. editor / M. Fikar ; M. Kvasnica. Institute of Electrical and Electronic Engineers IEEE, 2015. pp. 278-283
@inproceedings{395270386165425da267326969644669,
title = "Simple heuristic approach for training of type-2 NEO-fuzzy neural network",
abstract = "This paper describes the development of Interval Type-2 NEO-Fuzzy Neural Network for modeling of complex dynamics. The proposed network represents a parallel set of multiple zero order Sugeno type approximations, related only to their own input argument. As learning procedure a simple heuristic backpropagation approach, where the sign of the gradient is taken into account, is adopted. To improve the robustness of the network and the possibilities for handling uncertainties, Interval Type-2 Gaussian fuzzy sets are introduced into the network topology. The potentials of the proposed approach in modeling of Mackey-Glass and Rossler Chaotic time series are studied. A comparison is made with the classical Gradient Descent learning approach.",
keywords = "Chaotic time-series prediction, Dynamic modeling, Fuzzy systems, Neo-fuzzy neuron, Neural networks",
author = "Yancho Todorov and Margarita Terziyska",
year = "2015",
month = "7",
day = "28",
doi = "10.1109/PC.2015.7169976",
language = "English",
isbn = "978-1-4673-6627-4",
pages = "278--283",
editor = "M. Fikar and M. Kvasnica",
booktitle = "Proceedings of the 2015 20th International Conference on Process Control, PC 2015",
publisher = "Institute of Electrical and Electronic Engineers IEEE",
address = "United States",

}

Todorov, Y & Terziyska, M 2015, Simple heuristic approach for training of type-2 NEO-fuzzy neural network. in M Fikar & M Kvasnica (eds), Proceedings of the 2015 20th International Conference on Process Control, PC 2015., 7169976, Institute of Electrical and Electronic Engineers IEEE, pp. 278-283, 20th International Conference on Process Control, PC 2015, Strbske Pleso, Slovakia, 9/06/15. https://doi.org/10.1109/PC.2015.7169976

Simple heuristic approach for training of type-2 NEO-fuzzy neural network. / Todorov, Yancho; Terziyska, Margarita.

Proceedings of the 2015 20th International Conference on Process Control, PC 2015. ed. / M. Fikar; M. Kvasnica. Institute of Electrical and Electronic Engineers IEEE, 2015. p. 278-283 7169976.

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

TY - GEN

T1 - Simple heuristic approach for training of type-2 NEO-fuzzy neural network

AU - Todorov, Yancho

AU - Terziyska, Margarita

PY - 2015/7/28

Y1 - 2015/7/28

N2 - This paper describes the development of Interval Type-2 NEO-Fuzzy Neural Network for modeling of complex dynamics. The proposed network represents a parallel set of multiple zero order Sugeno type approximations, related only to their own input argument. As learning procedure a simple heuristic backpropagation approach, where the sign of the gradient is taken into account, is adopted. To improve the robustness of the network and the possibilities for handling uncertainties, Interval Type-2 Gaussian fuzzy sets are introduced into the network topology. The potentials of the proposed approach in modeling of Mackey-Glass and Rossler Chaotic time series are studied. A comparison is made with the classical Gradient Descent learning approach.

AB - This paper describes the development of Interval Type-2 NEO-Fuzzy Neural Network for modeling of complex dynamics. The proposed network represents a parallel set of multiple zero order Sugeno type approximations, related only to their own input argument. As learning procedure a simple heuristic backpropagation approach, where the sign of the gradient is taken into account, is adopted. To improve the robustness of the network and the possibilities for handling uncertainties, Interval Type-2 Gaussian fuzzy sets are introduced into the network topology. The potentials of the proposed approach in modeling of Mackey-Glass and Rossler Chaotic time series are studied. A comparison is made with the classical Gradient Descent learning approach.

KW - Chaotic time-series prediction

KW - Dynamic modeling

KW - Fuzzy systems

KW - Neo-fuzzy neuron

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=84990985649&partnerID=8YFLogxK

U2 - 10.1109/PC.2015.7169976

DO - 10.1109/PC.2015.7169976

M3 - Conference article in proceedings

AN - SCOPUS:84990985649

SN - 978-1-4673-6627-4

SP - 278

EP - 283

BT - Proceedings of the 2015 20th International Conference on Process Control, PC 2015

A2 - Fikar, M.

A2 - Kvasnica, M.

PB - Institute of Electrical and Electronic Engineers IEEE

ER -

Todorov Y, Terziyska M. Simple heuristic approach for training of type-2 NEO-fuzzy neural network. In Fikar M, Kvasnica M, editors, Proceedings of the 2015 20th International Conference on Process Control, PC 2015. Institute of Electrical and Electronic Engineers IEEE. 2015. p. 278-283. 7169976 https://doi.org/10.1109/PC.2015.7169976