Adaptive training and pruning in feedforward networks

Sheng-Jiang Chang; John Sum; Kwok-Wo Wong; Chi-Sing Leung

doi:10.1049/el:20010074

Adaptive training and pruning in feedforward networks

Sheng-Jiang Chang
, John Sum
, Kwok-Wo Wong
, Chi-Sing Leung

Department of Electrical Engineering

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

4 Citations (Scopus)

Abstract

A local extended Kalman filter training and pruning approach is proposed to train feedforward networks. This approach is proposed specifically to reduce the computational complexity and storage requirement in large-scale practical problems. The performance of the proposed algorithm is demonstrated for the problem of handwritten digit recognition.

Original language	English
Pages (from-to)	106-107
Journal	Electronics Letters
Volume	37
Issue number	2
DOIs	https://doi.org/10.1049/el:20010074
Publication status	Published - 18 Jan 2001

Access Output

10.1049/el:20010074

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{c5c6747e82b64cd5ba61e0edf0844004,

title = "Adaptive training and pruning in feedforward networks",

abstract = "A local extended Kalman filter training and pruning approach is proposed to train feedforward networks. This approach is proposed specifically to reduce the computational complexity and storage requirement in large-scale practical problems. The performance of the proposed algorithm is demonstrated for the problem of handwritten digit recognition.",

author = "Sheng-Jiang Chang and John Sum and Kwok-Wo Wong and Chi-Sing Leung",

year = "2001",

month = jan,

day = "18",

doi = "10.1049/el:20010074",

language = "English",

volume = "37",

pages = "106--107",

journal = "Electronics Letters",

issn = "0013-5194",

publisher = "John Wiley \& Sons",

number = "2",

}

TY - JOUR

T1 - Adaptive training and pruning in feedforward networks

AU - Chang, Sheng-Jiang

AU - Sum, John

AU - Wong, Kwok-Wo

AU - Leung, Chi-Sing

PY - 2001/1/18

Y1 - 2001/1/18

N2 - A local extended Kalman filter training and pruning approach is proposed to train feedforward networks. This approach is proposed specifically to reduce the computational complexity and storage requirement in large-scale practical problems. The performance of the proposed algorithm is demonstrated for the problem of handwritten digit recognition.

AB - A local extended Kalman filter training and pruning approach is proposed to train feedforward networks. This approach is proposed specifically to reduce the computational complexity and storage requirement in large-scale practical problems. The performance of the proposed algorithm is demonstrated for the problem of handwritten digit recognition.

UR - https://www.scopus.com/pages/publications/0035136577

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-0035136577&origin=recordpage

U2 - 10.1049/el:20010074

DO - 10.1049/el:20010074

M3 - RGC 21 - Publication in refereed journal

SN - 0013-5194

VL - 37

SP - 106

EP - 107

JO - Electronics Letters

JF - Electronics Letters

IS - 2

ER -

Adaptive training and pruning in feedforward networks

Abstract

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this