An on-line hybrid learning algorithm for multilayer perceptron in identification problems

Daohang Sha; Vladimir B. Baji

doi:10.1016/S0045-7906(00)00063-X

An on-line hybrid learning algorithm for multilayer perceptron in identification problems

Daohang Sha, Vladimir B. Baji

Research output: Contribution to journal › Article › peer-review

16 Scopus citations

Abstract

A hybrid learning algorithm for multilayered perceptrons (MLPs) and pattern-by-pattern training, based on optimized instantaneous learning rates and the recursive least squares method, is proposed. This hybrid solution is developed for on-line identification of process models based on the use of MLPs, and can speed up the learning process of the MLPs substantially, while simultaneously preserving the stability of the learning process. For illustration and test purposes the proposed algorithm is applied to the identification of a non-linear dynamic system.

Original language	English (US)
Pages (from-to)	587-598
Number of pages	12
Journal	Computers and Electrical Engineering
Volume	28
Issue number	6
DOIs	https://doi.org/10.1016/S0045-7906(00)00063-X
State	Published - Nov 2002
Externally published	Yes

Keywords

Gradient descent method
Hybrid learning
Neural networks
Recursive least squares parameter estimation
Soft computing

ASJC Scopus subject areas

Control and Systems Engineering
General Computer Science
Electrical and Electronic Engineering

Access to Document

10.1016/S0045-7906(00)00063-X

Cite this

@article{89ffcd66f985420b9404518d30109294,

title = "An on-line hybrid learning algorithm for multilayer perceptron in identification problems",

abstract = "A hybrid learning algorithm for multilayered perceptrons (MLPs) and pattern-by-pattern training, based on optimized instantaneous learning rates and the recursive least squares method, is proposed. This hybrid solution is developed for on-line identification of process models based on the use of MLPs, and can speed up the learning process of the MLPs substantially, while simultaneously preserving the stability of the learning process. For illustration and test purposes the proposed algorithm is applied to the identification of a non-linear dynamic system.",

keywords = "Gradient descent method, Hybrid learning, Neural networks, Recursive least squares parameter estimation, Soft computing",

author = "Daohang Sha and Baji, {Vladimir B.}",

year = "2002",

month = nov,

doi = "10.1016/S0045-7906(00)00063-X",

language = "English (US)",

volume = "28",

pages = "587--598",

journal = "Computers and Electrical Engineering",

issn = "0045-7906",

publisher = "Elsevier Ltd",

number = "6",

}

TY - JOUR

T1 - An on-line hybrid learning algorithm for multilayer perceptron in identification problems

AU - Sha, Daohang

AU - Baji, Vladimir B.

PY - 2002/11

Y1 - 2002/11

N2 - A hybrid learning algorithm for multilayered perceptrons (MLPs) and pattern-by-pattern training, based on optimized instantaneous learning rates and the recursive least squares method, is proposed. This hybrid solution is developed for on-line identification of process models based on the use of MLPs, and can speed up the learning process of the MLPs substantially, while simultaneously preserving the stability of the learning process. For illustration and test purposes the proposed algorithm is applied to the identification of a non-linear dynamic system.

AB - A hybrid learning algorithm for multilayered perceptrons (MLPs) and pattern-by-pattern training, based on optimized instantaneous learning rates and the recursive least squares method, is proposed. This hybrid solution is developed for on-line identification of process models based on the use of MLPs, and can speed up the learning process of the MLPs substantially, while simultaneously preserving the stability of the learning process. For illustration and test purposes the proposed algorithm is applied to the identification of a non-linear dynamic system.

KW - Gradient descent method

KW - Hybrid learning

KW - Neural networks

KW - Recursive least squares parameter estimation

KW - Soft computing

UR - http://www.scopus.com/inward/record.url?scp=0036832924&partnerID=8YFLogxK

U2 - 10.1016/S0045-7906(00)00063-X

DO - 10.1016/S0045-7906(00)00063-X

M3 - Article

AN - SCOPUS:0036832924

SN - 0045-7906

VL - 28

SP - 587

EP - 598

JO - Computers and Electrical Engineering

JF - Computers and Electrical Engineering

IS - 6

ER -

An on-line hybrid learning algorithm for multilayer perceptron in identification problems

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this