The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks

Joo Kim Kyoung, Bae Park Jin, Ho Choi Yoon

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

4 Citations (Scopus)

Abstract

Since the convergence of neural networks depends on learning rates, the learning rates of training algorithm for neural networks are very important factors. Therefore, we propose the Adaptive Learning Rates(ALRs) of Extended Kalman Filter(EKF) based training algorithm for wavelet neural networks(WNNs). The ALRs of the EFK based training algorithm produce the convergence of the WNN. Also we derive the convergence analysis of the learning process from the discrete Lyapunov stability theorem. Several simulation results show that the EKF based WNN with ALRs adapt to abrupt change and high nonlinearity with satisfactory performance.

Original languageEnglish
Title of host publicationMICAI 2006
Subtitle of host publicationAdvances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings
Pages327-337
Number of pages11
Publication statusPublished - 2006 Dec 1
Event5th Mexican International Conference on Artificial Intelligence, MICAI 2006: Advances in Artificial Intelligence - Apizaco, Mexico
Duration: 2006 Nov 132006 Nov 17

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4293 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other5th Mexican International Conference on Artificial Intelligence, MICAI 2006: Advances in Artificial Intelligence
CountryMexico
CityApizaco
Period06/11/1306/11/17

Fingerprint

Wavelet Neural Network
Adaptive Learning
Learning Rate
Training Algorithm
Extended Kalman filters
Kalman Filter
Learning
Neural networks
Neural Networks
Lyapunov Theorem
Lyapunov Stability
Stability Theorem
Learning Process
Convergence Analysis
Nonlinearity
Simulation

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Kyoung, J. K., Jin, B. P., & Yoon, H. C. (2006). The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks. In MICAI 2006: Advances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings (pp. 327-337). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4293 LNAI).
Kyoung, Joo Kim ; Jin, Bae Park ; Yoon, Ho Choi. / The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks. MICAI 2006: Advances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings. 2006. pp. 327-337 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{bce37db427a24b68b6373ab1a82bf7df,
title = "The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks",
abstract = "Since the convergence of neural networks depends on learning rates, the learning rates of training algorithm for neural networks are very important factors. Therefore, we propose the Adaptive Learning Rates(ALRs) of Extended Kalman Filter(EKF) based training algorithm for wavelet neural networks(WNNs). The ALRs of the EFK based training algorithm produce the convergence of the WNN. Also we derive the convergence analysis of the learning process from the discrete Lyapunov stability theorem. Several simulation results show that the EKF based WNN with ALRs adapt to abrupt change and high nonlinearity with satisfactory performance.",
author = "Kyoung, {Joo Kim} and Jin, {Bae Park} and Yoon, {Ho Choi}",
year = "2006",
month = "12",
day = "1",
language = "English",
isbn = "3540490264",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "327--337",
booktitle = "MICAI 2006",

}

Kyoung, JK, Jin, BP & Yoon, HC 2006, The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks. in MICAI 2006: Advances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4293 LNAI, pp. 327-337, 5th Mexican International Conference on Artificial Intelligence, MICAI 2006: Advances in Artificial Intelligence, Apizaco, Mexico, 06/11/13.

The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks. / Kyoung, Joo Kim; Jin, Bae Park; Yoon, Ho Choi.

MICAI 2006: Advances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings. 2006. p. 327-337 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4293 LNAI).

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

TY - GEN

T1 - The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks

AU - Kyoung, Joo Kim

AU - Jin, Bae Park

AU - Yoon, Ho Choi

PY - 2006/12/1

Y1 - 2006/12/1

N2 - Since the convergence of neural networks depends on learning rates, the learning rates of training algorithm for neural networks are very important factors. Therefore, we propose the Adaptive Learning Rates(ALRs) of Extended Kalman Filter(EKF) based training algorithm for wavelet neural networks(WNNs). The ALRs of the EFK based training algorithm produce the convergence of the WNN. Also we derive the convergence analysis of the learning process from the discrete Lyapunov stability theorem. Several simulation results show that the EKF based WNN with ALRs adapt to abrupt change and high nonlinearity with satisfactory performance.

AB - Since the convergence of neural networks depends on learning rates, the learning rates of training algorithm for neural networks are very important factors. Therefore, we propose the Adaptive Learning Rates(ALRs) of Extended Kalman Filter(EKF) based training algorithm for wavelet neural networks(WNNs). The ALRs of the EFK based training algorithm produce the convergence of the WNN. Also we derive the convergence analysis of the learning process from the discrete Lyapunov stability theorem. Several simulation results show that the EKF based WNN with ALRs adapt to abrupt change and high nonlinearity with satisfactory performance.

UR - http://www.scopus.com/inward/record.url?scp=33845930321&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845930321&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33845930321

SN - 3540490264

SN - 9783540490265

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 327

EP - 337

BT - MICAI 2006

ER -

Kyoung JK, Jin BP, Yoon HC. The adaptive learning rates of extended Kalman filter based training algorithm for wavelet neural networks. In MICAI 2006: Advances in Artificial Intelligence - 5th Mexican International Conference on Artificial Intelligence, Proceedings. 2006. p. 327-337. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).