Multi-branch neural networks with branch control

Takashi Yamashita, Kotaro Hirasawa, Takayuki Furuzuki

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

Abstract

Multi-branch neural networks have been proposed already in order to realize compact networks. It uses some branches between nodes, and this can improve the learning and generalization ability of the networks. In this paper, Branch Control is proposed on the multi-branch neural networks to further enhance the learning and generalization ability of the networks. Branch Control is to adjust the values of the signals on the branches depending on the network inputs using an additional branch control network. It has been clarified from simulation results of a function approximation problem that multi-branch neural networks with Branch Control could be improved more than that without Branch Control.

Original languageEnglish
Title of host publicationProceedings of the IEEE International Conference on Systems, Man and Cybernetics
Pages756-761
Number of pages6
Volume1
Publication statusPublished - 2003
EventSystem Security and Assurance - Washington, DC, United States
Duration: 2003 Oct 52003 Oct 8

Other

OtherSystem Security and Assurance
CountryUnited States
CityWashington, DC
Period03/10/503/10/8

Fingerprint

Neural networks

Keywords

  • Functional localization
  • Multi-branch
  • Neural networks

ASJC Scopus subject areas

  • Hardware and Architecture
  • Control and Systems Engineering

Cite this

Yamashita, T., Hirasawa, K., & Furuzuki, T. (2003). Multi-branch neural networks with branch control. In Proceedings of the IEEE International Conference on Systems, Man and Cybernetics (Vol. 1, pp. 756-761)

Multi-branch neural networks with branch control. / Yamashita, Takashi; Hirasawa, Kotaro; Furuzuki, Takayuki.

Proceedings of the IEEE International Conference on Systems, Man and Cybernetics. Vol. 1 2003. p. 756-761.

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

Yamashita, T, Hirasawa, K & Furuzuki, T 2003, Multi-branch neural networks with branch control. in Proceedings of the IEEE International Conference on Systems, Man and Cybernetics. vol. 1, pp. 756-761, System Security and Assurance, Washington, DC, United States, 03/10/5.
Yamashita T, Hirasawa K, Furuzuki T. Multi-branch neural networks with branch control. In Proceedings of the IEEE International Conference on Systems, Man and Cybernetics. Vol. 1. 2003. p. 756-761
Yamashita, Takashi ; Hirasawa, Kotaro ; Furuzuki, Takayuki. / Multi-branch neural networks with branch control. Proceedings of the IEEE International Conference on Systems, Man and Cybernetics. Vol. 1 2003. pp. 756-761
@inproceedings{fbe1f3a7ccd24679953128f46122fd3a,
title = "Multi-branch neural networks with branch control",
abstract = "Multi-branch neural networks have been proposed already in order to realize compact networks. It uses some branches between nodes, and this can improve the learning and generalization ability of the networks. In this paper, Branch Control is proposed on the multi-branch neural networks to further enhance the learning and generalization ability of the networks. Branch Control is to adjust the values of the signals on the branches depending on the network inputs using an additional branch control network. It has been clarified from simulation results of a function approximation problem that multi-branch neural networks with Branch Control could be improved more than that without Branch Control.",
keywords = "Functional localization, Multi-branch, Neural networks",
author = "Takashi Yamashita and Kotaro Hirasawa and Takayuki Furuzuki",
year = "2003",
language = "English",
volume = "1",
pages = "756--761",
booktitle = "Proceedings of the IEEE International Conference on Systems, Man and Cybernetics",

}

TY - GEN

T1 - Multi-branch neural networks with branch control

AU - Yamashita, Takashi

AU - Hirasawa, Kotaro

AU - Furuzuki, Takayuki

PY - 2003

Y1 - 2003

N2 - Multi-branch neural networks have been proposed already in order to realize compact networks. It uses some branches between nodes, and this can improve the learning and generalization ability of the networks. In this paper, Branch Control is proposed on the multi-branch neural networks to further enhance the learning and generalization ability of the networks. Branch Control is to adjust the values of the signals on the branches depending on the network inputs using an additional branch control network. It has been clarified from simulation results of a function approximation problem that multi-branch neural networks with Branch Control could be improved more than that without Branch Control.

AB - Multi-branch neural networks have been proposed already in order to realize compact networks. It uses some branches between nodes, and this can improve the learning and generalization ability of the networks. In this paper, Branch Control is proposed on the multi-branch neural networks to further enhance the learning and generalization ability of the networks. Branch Control is to adjust the values of the signals on the branches depending on the network inputs using an additional branch control network. It has been clarified from simulation results of a function approximation problem that multi-branch neural networks with Branch Control could be improved more than that without Branch Control.

KW - Functional localization

KW - Multi-branch

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=0242576463&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0242576463&partnerID=8YFLogxK

M3 - Conference contribution

VL - 1

SP - 756

EP - 761

BT - Proceedings of the IEEE International Conference on Systems, Man and Cybernetics

ER -