Back-propagation algorithm based on the extended Kalman filter

Aritoshi Kimura, Ikuo Arizono, Hiroshi Ohta

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Recently, Watanabe et al. proposed a back-propagation algorithm, in which the learning rate is time-varying, based on the extended Kalman filter (EKF). In the algorithm the interconnection strengths and biases are treated as the independent variables. However, the interconnection strengths and biases are not always independent, and have generally the mutual correlations. In this paper, we propose a new back-propagation algorithm in the case of considering the mutual correlations between the interconnection strengths and biases. Furthermore, by applying the proposed learning algorithm to the XOR problem and comparing with the algorithm of Watanabe et al., the ability of the proposed learning algorithm is examined.

Original languageEnglish
Title of host publicationProceedings of the International Joint Conference on Neural Networks
Editors Anon
PublisherPubl by IEEE
Pages1669-1672
Number of pages4
Volume2
ISBN (Print)0780314212
Publication statusPublished - 1993
Externally publishedYes
EventProceedings of 1993 International Joint Conference on Neural Networks. Part 2 (of 3) - Nagoya, Jpn
Duration: Oct 25 1993Oct 29 1993

Other

OtherProceedings of 1993 International Joint Conference on Neural Networks. Part 2 (of 3)
CityNagoya, Jpn
Period10/25/9310/29/93

Fingerprint

Backpropagation algorithms
Extended Kalman filters
Learning algorithms

ASJC Scopus subject areas

  • Software

Cite this

Kimura, A., Arizono, I., & Ohta, H. (1993). Back-propagation algorithm based on the extended Kalman filter. In Anon (Ed.), Proceedings of the International Joint Conference on Neural Networks (Vol. 2, pp. 1669-1672). Publ by IEEE.

Back-propagation algorithm based on the extended Kalman filter. / Kimura, Aritoshi; Arizono, Ikuo; Ohta, Hiroshi.

Proceedings of the International Joint Conference on Neural Networks. ed. / Anon. Vol. 2 Publ by IEEE, 1993. p. 1669-1672.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kimura, A, Arizono, I & Ohta, H 1993, Back-propagation algorithm based on the extended Kalman filter. in Anon (ed.), Proceedings of the International Joint Conference on Neural Networks. vol. 2, Publ by IEEE, pp. 1669-1672, Proceedings of 1993 International Joint Conference on Neural Networks. Part 2 (of 3), Nagoya, Jpn, 10/25/93.
Kimura A, Arizono I, Ohta H. Back-propagation algorithm based on the extended Kalman filter. In Anon, editor, Proceedings of the International Joint Conference on Neural Networks. Vol. 2. Publ by IEEE. 1993. p. 1669-1672
Kimura, Aritoshi ; Arizono, Ikuo ; Ohta, Hiroshi. / Back-propagation algorithm based on the extended Kalman filter. Proceedings of the International Joint Conference on Neural Networks. editor / Anon. Vol. 2 Publ by IEEE, 1993. pp. 1669-1672
@inproceedings{3943d14d2a8841b59d2396b79300c6fd,
title = "Back-propagation algorithm based on the extended Kalman filter",
abstract = "Recently, Watanabe et al. proposed a back-propagation algorithm, in which the learning rate is time-varying, based on the extended Kalman filter (EKF). In the algorithm the interconnection strengths and biases are treated as the independent variables. However, the interconnection strengths and biases are not always independent, and have generally the mutual correlations. In this paper, we propose a new back-propagation algorithm in the case of considering the mutual correlations between the interconnection strengths and biases. Furthermore, by applying the proposed learning algorithm to the XOR problem and comparing with the algorithm of Watanabe et al., the ability of the proposed learning algorithm is examined.",
author = "Aritoshi Kimura and Ikuo Arizono and Hiroshi Ohta",
year = "1993",
language = "English",
isbn = "0780314212",
volume = "2",
pages = "1669--1672",
editor = "Anon",
booktitle = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - Back-propagation algorithm based on the extended Kalman filter

AU - Kimura, Aritoshi

AU - Arizono, Ikuo

AU - Ohta, Hiroshi

PY - 1993

Y1 - 1993

N2 - Recently, Watanabe et al. proposed a back-propagation algorithm, in which the learning rate is time-varying, based on the extended Kalman filter (EKF). In the algorithm the interconnection strengths and biases are treated as the independent variables. However, the interconnection strengths and biases are not always independent, and have generally the mutual correlations. In this paper, we propose a new back-propagation algorithm in the case of considering the mutual correlations between the interconnection strengths and biases. Furthermore, by applying the proposed learning algorithm to the XOR problem and comparing with the algorithm of Watanabe et al., the ability of the proposed learning algorithm is examined.

AB - Recently, Watanabe et al. proposed a back-propagation algorithm, in which the learning rate is time-varying, based on the extended Kalman filter (EKF). In the algorithm the interconnection strengths and biases are treated as the independent variables. However, the interconnection strengths and biases are not always independent, and have generally the mutual correlations. In this paper, we propose a new back-propagation algorithm in the case of considering the mutual correlations between the interconnection strengths and biases. Furthermore, by applying the proposed learning algorithm to the XOR problem and comparing with the algorithm of Watanabe et al., the ability of the proposed learning algorithm is examined.

UR - http://www.scopus.com/inward/record.url?scp=0027857063&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027857063&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0780314212

VL - 2

SP - 1669

EP - 1672

BT - Proceedings of the International Joint Conference on Neural Networks

A2 - Anon, null

PB - Publ by IEEE

ER -