Hybrid training of recurrent fuzzy neural network model

Mojtaba Ahmadieh Khanesar; Mahdi Aliyari Shoorehdeli; Mohammad Teshnehlab

doi:10.1109/ICMA.2007.4303966

Hybrid training of recurrent fuzzy neural network model

Mojtaba Ahmadieh Khanesar^*, Mahdi Aliyari Shoorehdeli, Mohammad Teshnehlab

^*Corresponding author for this work

K.N. Toosi University of Technology

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

In this study, a hybrid learning algorithm for training the Recurrent Fuzzy Neural Network (RFNN) is introduced. This learning algorithm aims to solve main problems of the Gradient Descent (GD) based methods for the optimization of the RFNNs, which are instability, local minima and the problem of generalization of trained network to the test data. PSO as a global optimizer is used to optimize the parameters of the membership functions and the GD algorithm is used to optimize the consequent part's parameters of RFNN. As PSO is a derivative free optimization technique, a simpler method for the train of RFNN is achieved. Also the results are compared to GD algorithm.

Original language	English
Title of host publication	Proceedings of the 2007 IEEE International Conference on Mechatronics and Automation, ICMA 2007
Number of pages	6
Publication date	2007
Pages	2598-2603
Article number	4303966
ISBN (Print)	1424408288, 9781424408283
DOIs	https://doi.org/10.1109/ICMA.2007.4303966
Publication status	Published - 2007
Externally published	Yes
Event	2007 IEEE International Conference on Mechatronics and Automation - Harbin, China Duration: 5 Aug 2007 → 8 Aug 2007 https://ieeexplore.ieee.org/xpl/conhome/4303487/proceeding

Conference

Conference	2007 IEEE International Conference on Mechatronics and Automation
Country/Territory	China
City	Harbin
Period	05/08/2007 → 08/08/2007
Sponsor	IEEE, Harbin Engineering University, Kagawa University, National Natural Science Foundation of China, Ministry of Education, China
Internet address	https://ieeexplore.ieee.org/xpl/conhome/4303487/proceeding

Keywords

Gradient descent
Identification
Particle swarm optimization
Prediction
Recurrent fuzzy neural network

Access to Document

10.1109/ICMA.2007.4303966

OpenUrl availability

Full text

Cite this

@inproceedings{ebb8c096f0cb49ee81e6154e30f43d20,

title = "Hybrid training of recurrent fuzzy neural network model",

abstract = "In this study, a hybrid learning algorithm for training the Recurrent Fuzzy Neural Network (RFNN) is introduced. This learning algorithm aims to solve main problems of the Gradient Descent (GD) based methods for the optimization of the RFNNs, which are instability, local minima and the problem of generalization of trained network to the test data. PSO as a global optimizer is used to optimize the parameters of the membership functions and the GD algorithm is used to optimize the consequent part's parameters of RFNN. As PSO is a derivative free optimization technique, a simpler method for the train of RFNN is achieved. Also the results are compared to GD algorithm.",

keywords = "Gradient descent, Identification, Particle swarm optimization, Prediction, Recurrent fuzzy neural network",

author = "Khanesar, {Mojtaba Ahmadieh} and Shoorehdeli, {Mahdi Aliyari} and Mohammad Teshnehlab",

year = "2007",

doi = "10.1109/ICMA.2007.4303966",

language = "English",

isbn = "1424408288",

pages = "2598--2603",

booktitle = "Proceedings of the 2007 IEEE International Conference on Mechatronics and Automation, ICMA 2007",

note = "2007 IEEE International Conference on Mechatronics and Automation, ICMA 2007 ; Conference date: 05-08-2007 Through 08-08-2007",

url = "https://ieeexplore.ieee.org/xpl/conhome/4303487/proceeding",

}

Hybrid training of recurrent fuzzy neural network model. / Khanesar, Mojtaba Ahmadieh; Shoorehdeli, Mahdi Aliyari; Teshnehlab, Mohammad.
Proceedings of the 2007 IEEE International Conference on Mechatronics and Automation, ICMA 2007. 2007. p. 2598-2603 4303966.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Hybrid training of recurrent fuzzy neural network model

AU - Khanesar, Mojtaba Ahmadieh

AU - Shoorehdeli, Mahdi Aliyari

AU - Teshnehlab, Mohammad

PY - 2007

Y1 - 2007

N2 - In this study, a hybrid learning algorithm for training the Recurrent Fuzzy Neural Network (RFNN) is introduced. This learning algorithm aims to solve main problems of the Gradient Descent (GD) based methods for the optimization of the RFNNs, which are instability, local minima and the problem of generalization of trained network to the test data. PSO as a global optimizer is used to optimize the parameters of the membership functions and the GD algorithm is used to optimize the consequent part's parameters of RFNN. As PSO is a derivative free optimization technique, a simpler method for the train of RFNN is achieved. Also the results are compared to GD algorithm.

AB - In this study, a hybrid learning algorithm for training the Recurrent Fuzzy Neural Network (RFNN) is introduced. This learning algorithm aims to solve main problems of the Gradient Descent (GD) based methods for the optimization of the RFNNs, which are instability, local minima and the problem of generalization of trained network to the test data. PSO as a global optimizer is used to optimize the parameters of the membership functions and the GD algorithm is used to optimize the consequent part's parameters of RFNN. As PSO is a derivative free optimization technique, a simpler method for the train of RFNN is achieved. Also the results are compared to GD algorithm.

KW - Gradient descent

KW - Identification

KW - Particle swarm optimization

KW - Prediction

KW - Recurrent fuzzy neural network

U2 - 10.1109/ICMA.2007.4303966

DO - 10.1109/ICMA.2007.4303966

M3 - Article in proceedings

AN - SCOPUS:37049005142

SN - 1424408288

SN - 9781424408283

SP - 2598

EP - 2603

BT - Proceedings of the 2007 IEEE International Conference on Mechatronics and Automation, ICMA 2007

T2 - 2007 IEEE International Conference on Mechatronics and Automation

Y2 - 5 August 2007 through 8 August 2007

ER -

Hybrid training of recurrent fuzzy neural network model

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this