Generalized AdaTron algorithm

P. Nachbar; J. A. Nossek; J. Strobl

Generalized AdaTron algorithm

P. Nachbar, J. A. Nossek, J. Strobl

Chair of Circuit Theory and Signal Processing (2008 — 2021)

Technical University of Munich

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

8 Scopus citations

Abstract

The AdaTron algorithm is able to find the most insensitive weights of a perceptron or attractor network with respect to the usual euclidean norm. Firstly we define various notions of robustness for q-norms as appropriate for perceptrons or attractor networks. By extending the so called AdaTon theorem, we are able to generalize the AdaTron algorithm, which then finds the most insensitive weights with respect to an arbitrary q-norm (1<q<∞).

Original language	English
Title of host publication	Proceedings - IEEE International Symposium on Circuits and Systems
Publisher	Publ by IEEE
Pages	2152-2155
Number of pages	4
ISBN (Print)	0780312813
State	Published - 1993
Event	Proceedings of the 1993 IEEE International Symposium on Circuits and Systems - Chicago, IL, USA Duration: 3 May 1993 → 6 May 1993

Publication series

Name	Proceedings - IEEE International Symposium on Circuits and Systems
Volume	4
ISSN (Print)	0271-4310

Conference

Conference	Proceedings of the 1993 IEEE International Symposium on Circuits and Systems
City	Chicago, IL, USA
Period	3/05/93 → 6/05/93

Cite this

@inproceedings{4f6266965e764e0f85c31518bedf4036,

title = "Generalized AdaTron algorithm",

abstract = "The AdaTron algorithm is able to find the most insensitive weights of a perceptron or attractor network with respect to the usual euclidean norm. Firstly we define various notions of robustness for q-norms as appropriate for perceptrons or attractor networks. By extending the so called AdaTon theorem, we are able to generalize the AdaTron algorithm, which then finds the most insensitive weights with respect to an arbitrary q-norm (1",

author = "P. Nachbar and Nossek, {J. A.} and J. Strobl",

year = "1993",

language = "English",

isbn = "0780312813",

series = "Proceedings - IEEE International Symposium on Circuits and Systems",

publisher = "Publ by IEEE",

pages = "2152--2155",

booktitle = "Proceedings - IEEE International Symposium on Circuits and Systems",

note = "Proceedings of the 1993 IEEE International Symposium on Circuits and Systems ; Conference date: 03-05-1993 Through 06-05-1993",

}

TY - GEN

T1 - Generalized AdaTron algorithm

AU - Nachbar, P.

AU - Nossek, J. A.

AU - Strobl, J.

PY - 1993

Y1 - 1993

N2 - The AdaTron algorithm is able to find the most insensitive weights of a perceptron or attractor network with respect to the usual euclidean norm. Firstly we define various notions of robustness for q-norms as appropriate for perceptrons or attractor networks. By extending the so called AdaTon theorem, we are able to generalize the AdaTron algorithm, which then finds the most insensitive weights with respect to an arbitrary q-norm (1

AB - The AdaTron algorithm is able to find the most insensitive weights of a perceptron or attractor network with respect to the usual euclidean norm. Firstly we define various notions of robustness for q-norms as appropriate for perceptrons or attractor networks. By extending the so called AdaTon theorem, we are able to generalize the AdaTron algorithm, which then finds the most insensitive weights with respect to an arbitrary q-norm (1

UR - http://www.scopus.com/inward/record.url?scp=0027242793&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0027242793

SN - 0780312813

T3 - Proceedings - IEEE International Symposium on Circuits and Systems

SP - 2152

EP - 2155

BT - Proceedings - IEEE International Symposium on Circuits and Systems

PB - Publ by IEEE

T2 - Proceedings of the 1993 IEEE International Symposium on Circuits and Systems

Y2 - 3 May 1993 through 6 May 1993

ER -

Generalized AdaTron algorithm

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this