Self-organizing neural networks by construction and pruning

Jong Seok Lee, Hajoon Lee, Jae Young Kim, Dongkyung Nam, Cheol Hoon Park

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Feedforward neural networks have been successfully developed and applied in many areas because of their universal approximation capability. However, there still remains the problem of determining a suitable network structure for the given task. In this paper, we propose a novel self-organizing neural network which automatically adjusts its structure according to the task. Utilizing both the constructive and the pruning procedures, the proposed algorithm finds a near-optimal network which is compact and shows good generalization performance. One of its important features is reliability, which means the randomness of neural networks is effectively reduced. The resultant networks can have suitable numbers of hidden neurons and hidden layers according to the complexity of the given task. The simulation results for the well-known function regression problems show that our method successfully organizes near-optimal networks.

Original languageEnglish
Pages (from-to)2489-2498
Number of pages10
JournalIEICE Transactions on Information and Systems
VolumeE87-D
Issue number11
Publication statusPublished - 2004 Jan 1

Fingerprint

Neural networks
Feedforward neural networks
Neurons

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence
  • Electrical and Electronic Engineering

Cite this

Lee, J. S., Lee, H., Kim, J. Y., Nam, D., & Park, C. H. (2004). Self-organizing neural networks by construction and pruning. IEICE Transactions on Information and Systems, E87-D(11), 2489-2498.
Lee, Jong Seok ; Lee, Hajoon ; Kim, Jae Young ; Nam, Dongkyung ; Park, Cheol Hoon. / Self-organizing neural networks by construction and pruning. In: IEICE Transactions on Information and Systems. 2004 ; Vol. E87-D, No. 11. pp. 2489-2498.
@article{5c529848a2af4441bbc5bb34e7611169,
title = "Self-organizing neural networks by construction and pruning",
abstract = "Feedforward neural networks have been successfully developed and applied in many areas because of their universal approximation capability. However, there still remains the problem of determining a suitable network structure for the given task. In this paper, we propose a novel self-organizing neural network which automatically adjusts its structure according to the task. Utilizing both the constructive and the pruning procedures, the proposed algorithm finds a near-optimal network which is compact and shows good generalization performance. One of its important features is reliability, which means the randomness of neural networks is effectively reduced. The resultant networks can have suitable numbers of hidden neurons and hidden layers according to the complexity of the given task. The simulation results for the well-known function regression problems show that our method successfully organizes near-optimal networks.",
author = "Lee, {Jong Seok} and Hajoon Lee and Kim, {Jae Young} and Dongkyung Nam and Park, {Cheol Hoon}",
year = "2004",
month = "1",
day = "1",
language = "English",
volume = "E87-D",
pages = "2489--2498",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "11",

}

Lee, JS, Lee, H, Kim, JY, Nam, D & Park, CH 2004, 'Self-organizing neural networks by construction and pruning', IEICE Transactions on Information and Systems, vol. E87-D, no. 11, pp. 2489-2498.

Self-organizing neural networks by construction and pruning. / Lee, Jong Seok; Lee, Hajoon; Kim, Jae Young; Nam, Dongkyung; Park, Cheol Hoon.

In: IEICE Transactions on Information and Systems, Vol. E87-D, No. 11, 01.01.2004, p. 2489-2498.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Self-organizing neural networks by construction and pruning

AU - Lee, Jong Seok

AU - Lee, Hajoon

AU - Kim, Jae Young

AU - Nam, Dongkyung

AU - Park, Cheol Hoon

PY - 2004/1/1

Y1 - 2004/1/1

N2 - Feedforward neural networks have been successfully developed and applied in many areas because of their universal approximation capability. However, there still remains the problem of determining a suitable network structure for the given task. In this paper, we propose a novel self-organizing neural network which automatically adjusts its structure according to the task. Utilizing both the constructive and the pruning procedures, the proposed algorithm finds a near-optimal network which is compact and shows good generalization performance. One of its important features is reliability, which means the randomness of neural networks is effectively reduced. The resultant networks can have suitable numbers of hidden neurons and hidden layers according to the complexity of the given task. The simulation results for the well-known function regression problems show that our method successfully organizes near-optimal networks.

AB - Feedforward neural networks have been successfully developed and applied in many areas because of their universal approximation capability. However, there still remains the problem of determining a suitable network structure for the given task. In this paper, we propose a novel self-organizing neural network which automatically adjusts its structure according to the task. Utilizing both the constructive and the pruning procedures, the proposed algorithm finds a near-optimal network which is compact and shows good generalization performance. One of its important features is reliability, which means the randomness of neural networks is effectively reduced. The resultant networks can have suitable numbers of hidden neurons and hidden layers according to the complexity of the given task. The simulation results for the well-known function regression problems show that our method successfully organizes near-optimal networks.

UR - http://www.scopus.com/inward/record.url?scp=10444224637&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10444224637&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:10444224637

VL - E87-D

SP - 2489

EP - 2498

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 11

ER -