Partially connected neural networks for mapping problems

Sanggil Kang; Can Isik

Partially connected neural networks for mapping problems

Sanggil Kang, Can Isik

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Entry/Poem › Conference contribution

1 Scopus citations

Abstract

In this paper, we use partially connected feed-forward neural networks (PCFNNs) for input-output mapping problems to avoid a difficulty in determining epoch while fully connected feed-forward neural networks (FCFNNs) are being trained. PCFNNs can also, in some cases, improve generalization. Our method can be applicable to real input-output mapping problems such as blood pressure estimation and etc.

Original language	English (US)
Title of host publication	ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems
Editors	Slimane Hammoudi, Joaquim Filipe, Olivier Camp, Mario Piattini
Publisher	Escola Superior de Tecnologia do Instituto Politecnico de Setubal
Pages	469-473
Number of pages	5
ISBN (Electronic)	9729881618
State	Published - 2003
Event	5th International Conference on Enterprise Information Systems, ICEIS 2003 - Angers, France Duration: Apr 23 2003 → Apr 26 2003

Publication series

Name	ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems
Volume	2

Other

Other	5th International Conference on Enterprise Information Systems, ICEIS 2003
Country/Territory	France
City	Angers
Period	4/23/03 → 4/26/03

Keywords

Coupled input
Generalization
Input type
Partially connected feed-forward neural network
Uncoupled input

ASJC Scopus subject areas

Information Systems

Cite this

Kang, S., & Isik, C. (2003). Partially connected neural networks for mapping problems. In S. Hammoudi, J. Filipe, O. Camp, & M. Piattini (Eds.), ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems (pp. 469-473). (ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems; Vol. 2). Escola Superior de Tecnologia do Instituto Politecnico de Setubal.

Partially connected neural networks for mapping problems. / Kang, Sanggil; Isik, Can.
ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems. ed. / Slimane Hammoudi; Joaquim Filipe; Olivier Camp; Mario Piattini. Escola Superior de Tecnologia do Instituto Politecnico de Setubal, 2003. p. 469-473 (ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems; Vol. 2).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Kang, S & Isik, C 2003, Partially connected neural networks for mapping problems. in S Hammoudi, J Filipe, O Camp & M Piattini (eds), ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems. ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems, vol. 2, Escola Superior de Tecnologia do Instituto Politecnico de Setubal, pp. 469-473, 5th International Conference on Enterprise Information Systems, ICEIS 2003, Angers, France, 4/23/03.

Kang S, Isik C. Partially connected neural networks for mapping problems. In Hammoudi S, Filipe J, Camp O, Piattini M, editors, ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems. Escola Superior de Tecnologia do Instituto Politecnico de Setubal. 2003. p. 469-473. (ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems).

Kang, Sanggil ; Isik, Can. / Partially connected neural networks for mapping problems. ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems. editor / Slimane Hammoudi ; Joaquim Filipe ; Olivier Camp ; Mario Piattini. Escola Superior de Tecnologia do Instituto Politecnico de Setubal, 2003. pp. 469-473 (ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems).

@inproceedings{3e54111a2efe4872b8f362dc617724b1,

title = "Partially connected neural networks for mapping problems",

abstract = "In this paper, we use partially connected feed-forward neural networks (PCFNNs) for input-output mapping problems to avoid a difficulty in determining epoch while fully connected feed-forward neural networks (FCFNNs) are being trained. PCFNNs can also, in some cases, improve generalization. Our method can be applicable to real input-output mapping problems such as blood pressure estimation and etc.",

keywords = "Coupled input, Generalization, Input type, Partially connected feed-forward neural network, Uncoupled input",

author = "Sanggil Kang and Can Isik",

year = "2003",

language = "English (US)",

series = "ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems",

publisher = "Escola Superior de Tecnologia do Instituto Politecnico de Setubal",

pages = "469--473",

editor = "Slimane Hammoudi and Joaquim Filipe and Olivier Camp and Mario Piattini",

booktitle = "ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems",

}

TY - GEN

T1 - Partially connected neural networks for mapping problems

AU - Kang, Sanggil

AU - Isik, Can

PY - 2003

Y1 - 2003

N2 - In this paper, we use partially connected feed-forward neural networks (PCFNNs) for input-output mapping problems to avoid a difficulty in determining epoch while fully connected feed-forward neural networks (FCFNNs) are being trained. PCFNNs can also, in some cases, improve generalization. Our method can be applicable to real input-output mapping problems such as blood pressure estimation and etc.

AB - In this paper, we use partially connected feed-forward neural networks (PCFNNs) for input-output mapping problems to avoid a difficulty in determining epoch while fully connected feed-forward neural networks (FCFNNs) are being trained. PCFNNs can also, in some cases, improve generalization. Our method can be applicable to real input-output mapping problems such as blood pressure estimation and etc.

KW - Coupled input

KW - Generalization

KW - Input type

KW - Partially connected feed-forward neural network

KW - Uncoupled input

UR - http://www.scopus.com/inward/record.url?scp=22944465627&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=22944465627&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:22944465627

T3 - ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems

SP - 469

EP - 473

BT - ICEIS 2003 - Proceedings of the 5th International Conference on Enterprise Information Systems

A2 - Hammoudi, Slimane

A2 - Filipe, Joaquim

A2 - Camp, Olivier

A2 - Piattini, Mario

PB - Escola Superior de Tecnologia do Instituto Politecnico de Setubal

T2 - 5th International Conference on Enterprise Information Systems, ICEIS 2003

Y2 - 23 April 2003 through 26 April 2003

ER -

Partially connected neural networks for mapping problems

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this