Periodic activation function for fast on-line EKF training and pruning

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

7 Scopus Citations
Scopus Metrics
View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)2255-2257
Journal / PublicationElectronics Letters
Volume34
Issue number23
Publication statusPublished - 12 Nov 1998

Link(s)

DOI

DOI
Check@CityULib
Link to Scopus https://www.scopus.com/record/display.uri?eid=2-s2.0-0032206420&origin=recordpage
Permanent Link https://scholars.cityu.edu.hk/en/publications/publication(272cd799-d47b-4a15-98b0-242fcc0573d5).html

Abstract

In the use of the local extended Kalman filter algorithm for network training and pruning, the authors propose replacing the monotonical increasing sigmoid activation function with a periodic function. Computer simulations show that the rate of convergence is substantially improved and that more insignificant weights can be pruned.

Fingerprint

Publication fingerprints text

100
Convergence Rate Keyphrases
100
Pruning Keyphrases
100
Network Training Keyphrases
100
Periodic Function Keyphrases
100
Sigmoid Activation ... Keyphrases
100
Network Pruning Keyphrases
100
Periodic Activation... Keyphrases
100
Extended Kalman Fil... Keyphrases
100
algorithms INIS
100
convergence INIS
Full Fingerprint ›

Citation Format(s)

[ RIS ] [ BibTeX ]
Periodic activation function for fast on-line EKF training and pruning. / Chang, Sheng-Jiang; Wong, Kwok-Wo; Leung, Chi-Sing.
In: Electronics Letters, Vol. 34, No. 23, 12.11.1998, p. 2255-2257.

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review