Efficient neural network training algorithm for the Cray Y-MP supercomputer

Chung Siu Leung; Rudy Setiono

doi:10.1109/IJCNN.1993.717036

Efficient neural network training algorithm for the Cray Y-MP supercomputer

Chung Siu Leung
, Rudy Setiono

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

2 Citations (Scopus)

Abstract

An efficient implementation of a quasi-Newton algorithm for feedforward neural network training on a Cray Y-MP is presented. The most time-consuming step of a neural network training using the quasi-Newton algorithm is the computation of the error function and its gradient. We describe in this paper how this step can be implemented so that the neural network training may take full advantage of the Cray vectorization capabilities.

Original language	English
Title of host publication	IJCNN'93-NAGOYA
Subtitle of host publication	PROCEEDINGS OF 1993 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS
Publisher	IEEE
Pages	1943-1946
Volume	3
ISBN (Print)	0-7803-1421-2, 0-7803-1422-0
DOIs	https://doi.org/10.1109/IJCNN.1993.717036
Publication status	Published - Oct 1993
Externally published	Yes
Event	1993 International Joint Conference on Neural Networks (IJCNN'93-NAGOYA) - NAGOYA CONGRESS CENTER, Nagoya, Japan Duration: 25 Oct 1993 → 29 Oct 1993 https://ewh.ieee.org/conf/ijcnn/1993/ijcnn-1993Oct.pdf

Publication series

Name	International Joint Conference on Neural Networks (IJCNN)
Publisher	IEEE
Volume	1993

Conference

Conference	1993 International Joint Conference on Neural Networks (IJCNN'93-NAGOYA)
Place	Japan
City	Nagoya
Period	25/10/93 → 29/10/93
Internet address	https://ewh.ieee.org/conf/ijcnn/1993/ijcnn-1993Oct.pdf

Access Output

10.1109/IJCNN.1993.717036

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{b8b2f718f715476084b6e04b38cd1032,

title = "Efficient neural network training algorithm for the Cray Y-MP supercomputer",

abstract = "An efficient implementation of a quasi-Newton algorithm for feedforward neural network training on a Cray Y-MP is presented. The most time-consuming step of a neural network training using the quasi-Newton algorithm is the computation of the error function and its gradient. We describe in this paper how this step can be implemented so that the neural network training may take full advantage of the Cray vectorization capabilities.",

author = "Leung, \{Chung Siu\} and Rudy Setiono",

year = "1993",

month = oct,

doi = "10.1109/IJCNN.1993.717036",

language = "English",

isbn = "0-7803-1421-2",

volume = "3",

series = "International Joint Conference on Neural Networks (IJCNN) ",

publisher = "IEEE",

pages = "1943--1946",

booktitle = "IJCNN'93-NAGOYA",

address = "United States",

note = "1993 International Joint Conference on Neural Networks (IJCNN'93-NAGOYA) ; Conference date: 25-10-1993 Through 29-10-1993",

url = "https://ewh.ieee.org/conf/ijcnn/1993/ijcnn-1993Oct.pdf",

}

Leung, CS & Setiono, R 1993, Efficient neural network training algorithm for the Cray Y-MP supercomputer. in IJCNN'93-NAGOYA : PROCEEDINGS OF 1993 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS. vol. 3, International Joint Conference on Neural Networks (IJCNN) , vol. 1993, IEEE, pp. 1943-1946, 1993 International Joint Conference on Neural Networks (IJCNN'93-NAGOYA), Nagoya, Japan, 25/10/93. https://doi.org/10.1109/IJCNN.1993.717036

Efficient neural network training algorithm for the Cray Y-MP supercomputer. / Leung, Chung Siu; Setiono, Rudy.
IJCNN'93-NAGOYA : PROCEEDINGS OF 1993 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS. Vol. 3 IEEE, 1993. p. 1943-1946 (International Joint Conference on Neural Networks (IJCNN) ; Vol. 1993).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Efficient neural network training algorithm for the Cray Y-MP supercomputer

AU - Leung, Chung Siu

AU - Setiono, Rudy

PY - 1993/10

Y1 - 1993/10

N2 - An efficient implementation of a quasi-Newton algorithm for feedforward neural network training on a Cray Y-MP is presented. The most time-consuming step of a neural network training using the quasi-Newton algorithm is the computation of the error function and its gradient. We describe in this paper how this step can be implemented so that the neural network training may take full advantage of the Cray vectorization capabilities.

AB - An efficient implementation of a quasi-Newton algorithm for feedforward neural network training on a Cray Y-MP is presented. The most time-consuming step of a neural network training using the quasi-Newton algorithm is the computation of the error function and its gradient. We describe in this paper how this step can be implemented so that the neural network training may take full advantage of the Cray vectorization capabilities.

UR - https://www.scopus.com/pages/publications/0027845153

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-0027845153&origin=recordpage

U2 - 10.1109/IJCNN.1993.717036

DO - 10.1109/IJCNN.1993.717036

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 0-7803-1421-2

SN - 0-7803-1422-0

VL - 3

T3 - International Joint Conference on Neural Networks (IJCNN)

SP - 1943

EP - 1946

BT - IJCNN'93-NAGOYA

PB - IEEE

T2 - 1993 International Joint Conference on Neural Networks (IJCNN'93-NAGOYA)

Y2 - 25 October 1993 through 29 October 1993

ER -

Efficient neural network training algorithm for the Cray Y-MP supercomputer

Abstract

Publication series

Conference

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this