Publication detail

Stability of backpropagation learning rule

KRUPANSKÝ, P., PIVOŇKA, P.

Original Title

Stability of backpropagation learning rule

Type

conference paper

Language

English

Original Abstract

A control of real processes requires different approach to neural network learning. The presented modification of backpropagation learning algorithm changes a meaning of learning constants. A base of modification is stability condition of learning dynamics.

Keywords

Neural networks, ARMA model, control, back-propagation, stability, the largest singular value, Euclidean norm

Authors

KRUPANSKÝ, P., PIVOŇKA, P.

RIV year

2003

Released

15. 9. 2003

Publisher

Rektor der Hochschule Zittau/Gorlitz

Location

Zittau, Germany

ISBN

3-9808089-4-7

Book

Third EUSFLAT Proceedings

Edition

Neuveden

Edition number

1

Pages from

369

Pages to

374

Pages count

6

BibTex

@inproceedings{BUT8168,
  author="Petr {Krupanský} and Petr {Pivoňka}",
  title="Stability of backpropagation learning rule",
  booktitle="Third EUSFLAT Proceedings",
  year="2003",
  series="Neuveden",
  volume="Neuveden",
  number="1",
  pages="6",
  publisher="Rektor der Hochschule Zittau/Gorlitz",
  address="Zittau, Germany",
  isbn="3-9808089-4-7"
}