Supervised learning with artificial selection

Masafumi Hagiwara, Masao Nakagawa

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

4 Citations (Scopus)

Abstract

Summary form only given, as follows. Supervised learning with artificial selection is proposed as a way to escape from local minima. The concept of artificial selection is reasonable for nature. In the authors' method, the 'worst' hidden unit is detected, and then all the weights connected to the detected hidden unit are reset to small random values. According to simulations, only half the trials using conventional backpropagation converge, whereas all of the trials using the proposed method converge, and quickly do so.

Original languageEnglish
Title of host publicationIJCNN Int Jt Conf Neural Network
Editors Anon
PublisherPubl by IEEE
Pages611
Number of pages1
Publication statusPublished - 1989
EventIJCNN International Joint Conference on Neural Networks - Washington, DC, USA
Duration: 1989 Jun 181989 Jun 22

Other

OtherIJCNN International Joint Conference on Neural Networks
CityWashington, DC, USA
Period89/6/1889/6/22

Fingerprint

Supervised learning
Backpropagation

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Hagiwara, M., & Nakagawa, M. (1989). Supervised learning with artificial selection. In Anon (Ed.), IJCNN Int Jt Conf Neural Network (pp. 611). Publ by IEEE.

Supervised learning with artificial selection. / Hagiwara, Masafumi; Nakagawa, Masao.

IJCNN Int Jt Conf Neural Network. ed. / Anon. Publ by IEEE, 1989. p. 611.

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

Hagiwara, M & Nakagawa, M 1989, Supervised learning with artificial selection. in Anon (ed.), IJCNN Int Jt Conf Neural Network. Publ by IEEE, pp. 611, IJCNN International Joint Conference on Neural Networks, Washington, DC, USA, 89/6/18.
Hagiwara M, Nakagawa M. Supervised learning with artificial selection. In Anon, editor, IJCNN Int Jt Conf Neural Network. Publ by IEEE. 1989. p. 611
Hagiwara, Masafumi ; Nakagawa, Masao. / Supervised learning with artificial selection. IJCNN Int Jt Conf Neural Network. editor / Anon. Publ by IEEE, 1989. pp. 611
@inproceedings{b46e64ce8caf4c88b5d1854c6a896998,
title = "Supervised learning with artificial selection",
abstract = "Summary form only given, as follows. Supervised learning with artificial selection is proposed as a way to escape from local minima. The concept of artificial selection is reasonable for nature. In the authors' method, the 'worst' hidden unit is detected, and then all the weights connected to the detected hidden unit are reset to small random values. According to simulations, only half the trials using conventional backpropagation converge, whereas all of the trials using the proposed method converge, and quickly do so.",
author = "Masafumi Hagiwara and Masao Nakagawa",
year = "1989",
language = "English",
pages = "611",
editor = "Anon",
booktitle = "IJCNN Int Jt Conf Neural Network",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - Supervised learning with artificial selection

AU - Hagiwara, Masafumi

AU - Nakagawa, Masao

PY - 1989

Y1 - 1989

N2 - Summary form only given, as follows. Supervised learning with artificial selection is proposed as a way to escape from local minima. The concept of artificial selection is reasonable for nature. In the authors' method, the 'worst' hidden unit is detected, and then all the weights connected to the detected hidden unit are reset to small random values. According to simulations, only half the trials using conventional backpropagation converge, whereas all of the trials using the proposed method converge, and quickly do so.

AB - Summary form only given, as follows. Supervised learning with artificial selection is proposed as a way to escape from local minima. The concept of artificial selection is reasonable for nature. In the authors' method, the 'worst' hidden unit is detected, and then all the weights connected to the detected hidden unit are reset to small random values. According to simulations, only half the trials using conventional backpropagation converge, whereas all of the trials using the proposed method converge, and quickly do so.

UR - http://www.scopus.com/inward/record.url?scp=0024879678&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024879678&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0024879678

SP - 611

BT - IJCNN Int Jt Conf Neural Network

A2 - Anon, null

PB - Publ by IEEE

ER -