The classification capabilities of exact two-layered perceptrons

P.J. Zwietering, E.H.L. Aarts, J. Wessels

Research output: Book/ReportBookScientific

Abstract

The present paper studies the problem of finding a two-layered perceptron that exactly classifies a given subset. Such a two-layered perceptron is called exact with respect to the given subset. We derive both a necessary and a sufficient condition for a given subset to be classifiable by an exact two-layered perceptron. The necessary condition can be viewed as a generalization of the linear-seperability condition of the one-layered perceptron and confirms the conjecture that the capabilities of exact two-layered perceptrons are more limited than those of exact three-layered perceptrons. The sufficient condition shows that the capabilities of exact two-layered perceptrons extend beyond the exact classification of convex subsets. Furthermore, we present a systematic verification method for the given sufficient condition. Keywords: Classification, Multi-Layered Perceptrons,Size Hidden Layer.
Original languageEnglish
Place of PublicationEindhoven
PublisherTechnische Universiteit Eindhoven
Publication statusPublished - 1991
Externally publishedYes

Publication series

NameMemorandum COSOR

Fingerprint

Neural networks

Cite this

Zwietering, P. J., Aarts, E. H. L., & Wessels, J. (1991). The classification capabilities of exact two-layered perceptrons. (Memorandum COSOR). Eindhoven: Technische Universiteit Eindhoven.
Zwietering, P.J. ; Aarts, E.H.L. ; Wessels, J. / The classification capabilities of exact two-layered perceptrons. Eindhoven : Technische Universiteit Eindhoven, 1991. (Memorandum COSOR).
@book{c742b12c1ff742b981e428c5f24c1f36,
title = "The classification capabilities of exact two-layered perceptrons",
abstract = "The present paper studies the problem of finding a two-layered perceptron that exactly classifies a given subset. Such a two-layered perceptron is called exact with respect to the given subset. We derive both a necessary and a sufficient condition for a given subset to be classifiable by an exact two-layered perceptron. The necessary condition can be viewed as a generalization of the linear-seperability condition of the one-layered perceptron and confirms the conjecture that the capabilities of exact two-layered perceptrons are more limited than those of exact three-layered perceptrons. The sufficient condition shows that the capabilities of exact two-layered perceptrons extend beyond the exact classification of convex subsets. Furthermore, we present a systematic verification method for the given sufficient condition. Keywords: Classification, Multi-Layered Perceptrons,Size Hidden Layer.",
author = "P.J. Zwietering and E.H.L. Aarts and J. Wessels",
year = "1991",
language = "English",
series = "Memorandum COSOR",
publisher = "Technische Universiteit Eindhoven",

}

Zwietering, PJ, Aarts, EHL & Wessels, J 1991, The classification capabilities of exact two-layered perceptrons. Memorandum COSOR, Technische Universiteit Eindhoven, Eindhoven.

The classification capabilities of exact two-layered perceptrons. / Zwietering, P.J.; Aarts, E.H.L.; Wessels, J.

Eindhoven : Technische Universiteit Eindhoven, 1991. (Memorandum COSOR).

Research output: Book/ReportBookScientific

TY - BOOK

T1 - The classification capabilities of exact two-layered perceptrons

AU - Zwietering, P.J.

AU - Aarts, E.H.L.

AU - Wessels, J.

PY - 1991

Y1 - 1991

N2 - The present paper studies the problem of finding a two-layered perceptron that exactly classifies a given subset. Such a two-layered perceptron is called exact with respect to the given subset. We derive both a necessary and a sufficient condition for a given subset to be classifiable by an exact two-layered perceptron. The necessary condition can be viewed as a generalization of the linear-seperability condition of the one-layered perceptron and confirms the conjecture that the capabilities of exact two-layered perceptrons are more limited than those of exact three-layered perceptrons. The sufficient condition shows that the capabilities of exact two-layered perceptrons extend beyond the exact classification of convex subsets. Furthermore, we present a systematic verification method for the given sufficient condition. Keywords: Classification, Multi-Layered Perceptrons,Size Hidden Layer.

AB - The present paper studies the problem of finding a two-layered perceptron that exactly classifies a given subset. Such a two-layered perceptron is called exact with respect to the given subset. We derive both a necessary and a sufficient condition for a given subset to be classifiable by an exact two-layered perceptron. The necessary condition can be viewed as a generalization of the linear-seperability condition of the one-layered perceptron and confirms the conjecture that the capabilities of exact two-layered perceptrons are more limited than those of exact three-layered perceptrons. The sufficient condition shows that the capabilities of exact two-layered perceptrons extend beyond the exact classification of convex subsets. Furthermore, we present a systematic verification method for the given sufficient condition. Keywords: Classification, Multi-Layered Perceptrons,Size Hidden Layer.

M3 - Book

T3 - Memorandum COSOR

BT - The classification capabilities of exact two-layered perceptrons

PB - Technische Universiteit Eindhoven

CY - Eindhoven

ER -

Zwietering PJ, Aarts EHL, Wessels J. The classification capabilities of exact two-layered perceptrons. Eindhoven: Technische Universiteit Eindhoven, 1991. (Memorandum COSOR).