Function Approximation Using Fuzzy Neural Networks with Robust Learning Algorithm

Wei Yen Wang, Tsu Tian Lee, Ching Lang Liu, Chi Hsu Wang

Research output: Contribution to journalArticle

94 Citations (Scopus)

Abstract

The paper describes a novel application of the B-spline membership functions (BMF's) and the fuzzy neural network to the function approximation with outliers in training data. According to the robust objective function, we use gradient descent method to derive the new learning rules of the weighting values and BMF's of the fuzzy neural network for robust function approximation. In this paper, the robust learning algorithm is derived. During the learning process, the robust objective function comes into effect and the approximated function will gradually be unaffected by the erroneous training data. As a result, the robust function approximation can rapidly converge to the desired tolerable error scope. In other words, the learning iterations will decrease greatly. We realize the function approximation not only in one dimension (curves), but also in two dimension (surfaces). Several examples are simulated in order to confirm the efficiency and feasibility of the proposed approach in this paper.

Original languageEnglish
Pages (from-to)740-747
Number of pages8
JournalIEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Volume27
Issue number4
DOIs
Publication statusPublished - 1997 Aug

Fingerprint

Fuzzy neural networks
Learning algorithms
Learning
Membership functions
Splines
Efficiency

Keywords

  • Function approximation
  • fuzzy neural network
  • robust learning algorithm

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Function Approximation Using Fuzzy Neural Networks with Robust Learning Algorithm. / Wang, Wei Yen; Lee, Tsu Tian; Liu, Ching Lang; Wang, Chi Hsu.

In: IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, Vol. 27, No. 4, 08.1997, p. 740-747.

Research output: Contribution to journalArticle

@article{bd9e19248277444791f8f6c456db266a,
title = "Function Approximation Using Fuzzy Neural Networks with Robust Learning Algorithm",
abstract = "The paper describes a novel application of the B-spline membership functions (BMF's) and the fuzzy neural network to the function approximation with outliers in training data. According to the robust objective function, we use gradient descent method to derive the new learning rules of the weighting values and BMF's of the fuzzy neural network for robust function approximation. In this paper, the robust learning algorithm is derived. During the learning process, the robust objective function comes into effect and the approximated function will gradually be unaffected by the erroneous training data. As a result, the robust function approximation can rapidly converge to the desired tolerable error scope. In other words, the learning iterations will decrease greatly. We realize the function approximation not only in one dimension (curves), but also in two dimension (surfaces). Several examples are simulated in order to confirm the efficiency and feasibility of the proposed approach in this paper.",
keywords = "Function approximation, fuzzy neural network, robust learning algorithm",
author = "Wang, {Wei Yen} and Lee, {Tsu Tian} and Liu, {Ching Lang} and Wang, {Chi Hsu}",
year = "1997",
month = "8",
doi = "10.1109/3477.604123",
language = "English",
volume = "27",
pages = "740--747",
journal = "IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics",
issn = "1083-4419",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Function Approximation Using Fuzzy Neural Networks with Robust Learning Algorithm

AU - Wang, Wei Yen

AU - Lee, Tsu Tian

AU - Liu, Ching Lang

AU - Wang, Chi Hsu

PY - 1997/8

Y1 - 1997/8

N2 - The paper describes a novel application of the B-spline membership functions (BMF's) and the fuzzy neural network to the function approximation with outliers in training data. According to the robust objective function, we use gradient descent method to derive the new learning rules of the weighting values and BMF's of the fuzzy neural network for robust function approximation. In this paper, the robust learning algorithm is derived. During the learning process, the robust objective function comes into effect and the approximated function will gradually be unaffected by the erroneous training data. As a result, the robust function approximation can rapidly converge to the desired tolerable error scope. In other words, the learning iterations will decrease greatly. We realize the function approximation not only in one dimension (curves), but also in two dimension (surfaces). Several examples are simulated in order to confirm the efficiency and feasibility of the proposed approach in this paper.

AB - The paper describes a novel application of the B-spline membership functions (BMF's) and the fuzzy neural network to the function approximation with outliers in training data. According to the robust objective function, we use gradient descent method to derive the new learning rules of the weighting values and BMF's of the fuzzy neural network for robust function approximation. In this paper, the robust learning algorithm is derived. During the learning process, the robust objective function comes into effect and the approximated function will gradually be unaffected by the erroneous training data. As a result, the robust function approximation can rapidly converge to the desired tolerable error scope. In other words, the learning iterations will decrease greatly. We realize the function approximation not only in one dimension (curves), but also in two dimension (surfaces). Several examples are simulated in order to confirm the efficiency and feasibility of the proposed approach in this paper.

KW - Function approximation

KW - fuzzy neural network

KW - robust learning algorithm

UR - http://www.scopus.com/inward/record.url?scp=0031212001&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031212001&partnerID=8YFLogxK

U2 - 10.1109/3477.604123

DO - 10.1109/3477.604123

M3 - Article

AN - SCOPUS:0031212001

VL - 27

SP - 740

EP - 747

JO - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

JF - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

SN - 1083-4419

IS - 4

ER -