Continuation-based learning algorithm for discrete-time cellular neural networks

Holger Magnussen; Georgios Papoutsis; Josef A. Nossek

Continuation-based learning algorithm for discrete-time cellular neural networks

Holger Magnussen, Georgios Papoutsis, Josef A. Nossek

Lehrstuhl für Netzwerktheorie und Signalverarbeitung (2008 — 2021)

TUM - Lehrstuhl für Architekturinformatik, Prof. Dr.-Ing. Frank Petzold

Publikation: Konferenzbeitrag › Papier › Begutachtung

5 Zitate (Scopus)

Abstract

The SGN-type nonlinearity of a standard Discrete-Time Cellular Neural Network (DTCNN) is replaced by a smooth, sigmoidal nonlinearity with variable gain. Therefore, the resulting dynamical system is fully differentiable. Bounds on gain of the sigmoidal function are given, so that the new, smooth system approximates the standard DTCNN within certain limits. A learning algorithm is proposed, which finds the template parameters for the standard DTCNN by gradually increasing the gain of the sigmoidal function.

Originalsprache	Englisch
Seiten	171-176
Seitenumfang	6
Publikationsstatus	Veröffentlicht - 1994
Veranstaltung	Proceedings of the 3rd IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94) - Rome, Italy Dauer: 18 Dez. 1994 → 21 Dez. 1994

Konferenz

Konferenz	Proceedings of the 3rd IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)
Ort	Rome, Italy
Zeitraum	18/12/94 → 21/12/94

Andere Dateien und Links

Verknüpfung zur Publikation in Scopus

Dieses zitieren

@conference{7aa24e31d72d428d92f4b0af1a6280b9,

title = "Continuation-based learning algorithm for discrete-time cellular neural networks",

abstract = "The SGN-type nonlinearity of a standard Discrete-Time Cellular Neural Network (DTCNN) is replaced by a smooth, sigmoidal nonlinearity with variable gain. Therefore, the resulting dynamical system is fully differentiable. Bounds on gain of the sigmoidal function are given, so that the new, smooth system approximates the standard DTCNN within certain limits. A learning algorithm is proposed, which finds the template parameters for the standard DTCNN by gradually increasing the gain of the sigmoidal function.",

author = "Holger Magnussen and Georgios Papoutsis and Nossek, {Josef A.}",

year = "1994",

language = "English",

pages = "171--176",

note = "Proceedings of the 3rd IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94) ; Conference date: 18-12-1994 Through 21-12-1994",

}

TY - CONF

T1 - Continuation-based learning algorithm for discrete-time cellular neural networks

AU - Magnussen, Holger

AU - Papoutsis, Georgios

AU - Nossek, Josef A.

PY - 1994

Y1 - 1994

N2 - The SGN-type nonlinearity of a standard Discrete-Time Cellular Neural Network (DTCNN) is replaced by a smooth, sigmoidal nonlinearity with variable gain. Therefore, the resulting dynamical system is fully differentiable. Bounds on gain of the sigmoidal function are given, so that the new, smooth system approximates the standard DTCNN within certain limits. A learning algorithm is proposed, which finds the template parameters for the standard DTCNN by gradually increasing the gain of the sigmoidal function.

AB - The SGN-type nonlinearity of a standard Discrete-Time Cellular Neural Network (DTCNN) is replaced by a smooth, sigmoidal nonlinearity with variable gain. Therefore, the resulting dynamical system is fully differentiable. Bounds on gain of the sigmoidal function are given, so that the new, smooth system approximates the standard DTCNN within certain limits. A learning algorithm is proposed, which finds the template parameters for the standard DTCNN by gradually increasing the gain of the sigmoidal function.

UR - http://www.scopus.com/inward/record.url?scp=0028749516&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0028749516

SP - 171

EP - 176

T2 - Proceedings of the 3rd IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)

Y2 - 18 December 1994 through 21 December 1994

ER -

Continuation-based learning algorithm for discrete-time cellular neural networks

Abstract

Konferenz

Andere Dateien und Links

Fingerprint

Dieses zitieren