Constrained Learning in Neural Networks

Application to Stable Factorization of 2-D Polynomials

Stavros Perantonis, Nikolaos Ampazis, Stavros Varoufakis, George Antoniou

Research output: Contribution to journalArticleResearchpeer-review

23 Citations (Scopus)

Abstract

Adaptive artificial neural network techniques are introduced and applied to the factorization of 2-D second order polynomials. The proposed neural network is trained using a constrained learning algorithm that achieves minimization of the usual mean square error criterion along with simultaneous satisfaction of multiple equality and inequality constraints between the polynomial coefficients. Using this method, we are able to obtain good approximate solutions for non-factorable polynomials. By incorporating stability constraints into the formalism, our method can be successfully used for the realization of stable 2-D second order IIR filters in cascade form.

Original languageEnglish
Pages (from-to)5-14
Number of pages10
JournalNeural Processing Letters
Volume7
Issue number1
DOIs
StatePublished - 1 Jan 1998

Fingerprint

Factorization
Polynomials
Learning
Neural networks
IIR filters
Mean square error
Learning algorithms

Keywords

  • Constrained learning
  • Factorization
  • Feedforward networks
  • IIR filters
  • Polynomials
  • Stability

Cite this

Perantonis, Stavros ; Ampazis, Nikolaos ; Varoufakis, Stavros ; Antoniou, George. / Constrained Learning in Neural Networks : Application to Stable Factorization of 2-D Polynomials. In: Neural Processing Letters. 1998 ; Vol. 7, No. 1. pp. 5-14.
@article{c31e258d6b52409ebdd5e18cce0ad176,
title = "Constrained Learning in Neural Networks: Application to Stable Factorization of 2-D Polynomials",
abstract = "Adaptive artificial neural network techniques are introduced and applied to the factorization of 2-D second order polynomials. The proposed neural network is trained using a constrained learning algorithm that achieves minimization of the usual mean square error criterion along with simultaneous satisfaction of multiple equality and inequality constraints between the polynomial coefficients. Using this method, we are able to obtain good approximate solutions for non-factorable polynomials. By incorporating stability constraints into the formalism, our method can be successfully used for the realization of stable 2-D second order IIR filters in cascade form.",
keywords = "Constrained learning, Factorization, Feedforward networks, IIR filters, Polynomials, Stability",
author = "Stavros Perantonis and Nikolaos Ampazis and Stavros Varoufakis and George Antoniou",
year = "1998",
month = "1",
day = "1",
doi = "10.1023/A:1009655902122",
language = "English",
volume = "7",
pages = "5--14",
journal = "Neural Processing Letters",
issn = "1370-4621",
publisher = "Springer Netherlands",
number = "1",

}

Constrained Learning in Neural Networks : Application to Stable Factorization of 2-D Polynomials. / Perantonis, Stavros; Ampazis, Nikolaos; Varoufakis, Stavros; Antoniou, George.

In: Neural Processing Letters, Vol. 7, No. 1, 01.01.1998, p. 5-14.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Constrained Learning in Neural Networks

T2 - Application to Stable Factorization of 2-D Polynomials

AU - Perantonis, Stavros

AU - Ampazis, Nikolaos

AU - Varoufakis, Stavros

AU - Antoniou, George

PY - 1998/1/1

Y1 - 1998/1/1

N2 - Adaptive artificial neural network techniques are introduced and applied to the factorization of 2-D second order polynomials. The proposed neural network is trained using a constrained learning algorithm that achieves minimization of the usual mean square error criterion along with simultaneous satisfaction of multiple equality and inequality constraints between the polynomial coefficients. Using this method, we are able to obtain good approximate solutions for non-factorable polynomials. By incorporating stability constraints into the formalism, our method can be successfully used for the realization of stable 2-D second order IIR filters in cascade form.

AB - Adaptive artificial neural network techniques are introduced and applied to the factorization of 2-D second order polynomials. The proposed neural network is trained using a constrained learning algorithm that achieves minimization of the usual mean square error criterion along with simultaneous satisfaction of multiple equality and inequality constraints between the polynomial coefficients. Using this method, we are able to obtain good approximate solutions for non-factorable polynomials. By incorporating stability constraints into the formalism, our method can be successfully used for the realization of stable 2-D second order IIR filters in cascade form.

KW - Constrained learning

KW - Factorization

KW - Feedforward networks

KW - IIR filters

KW - Polynomials

KW - Stability

UR - http://www.scopus.com/inward/record.url?scp=0031990532&partnerID=8YFLogxK

U2 - 10.1023/A:1009655902122

DO - 10.1023/A:1009655902122

M3 - Article

VL - 7

SP - 5

EP - 14

JO - Neural Processing Letters

JF - Neural Processing Letters

SN - 1370-4621

IS - 1

ER -