Bounds on the Number of Samples Needed for Neural Learning

Kishan G. Mehrotra; Chilukuri K. Mohan; Sanjay Ranka

doi:10.1109/72.97932

Bounds on the Number of Samples Needed for Neural Learning

Kishan G. Mehrotra, Chilukuri K. Mohan, Sanjay Ranka

Department of Electrical Engineering & Computer Science

Research output: Contribution to journal › Article › peer-review

66 Scopus citations

Abstract

This paper addresses the relationship between the number of hidden layer nodes in a neural network, the complexity of a multiclass discrimination problem, and the number of samples needed for effective learning. Bounds are given for the latter. We show that Ω(min (d, n). M) boundary samples are required for successful classification of M clusters of samples using a two-hidden-layer neural network with d-dimensional inputs and n nodes in the first hidden layer.

Original language	English (US)
Pages (from-to)	548-558
Number of pages	11
Journal	IEEE Transactions on Neural Networks
Volume	2
Issue number	6
DOIs	https://doi.org/10.1109/72.97932
State	Published - Nov 1991

ASJC Scopus subject areas

Software
Computer Science Applications
Computer Networks and Communications
Artificial Intelligence

Access to Document

10.1109/72.97932

Cite this

@article{a0d1141575134105842fbe8f9bdb9737,

title = "Bounds on the Number of Samples Needed for Neural Learning",

abstract = "This paper addresses the relationship between the number of hidden layer nodes in a neural network, the complexity of a multiclass discrimination problem, and the number of samples needed for effective learning. Bounds are given for the latter. We show that Ω(min (d, n). M) boundary samples are required for successful classification of M clusters of samples using a two-hidden-layer neural network with d-dimensional inputs and n nodes in the first hidden layer.",

author = "Mehrotra, {Kishan G.} and Mohan, {Chilukuri K.} and Sanjay Ranka",

year = "1991",

month = nov,

doi = "10.1109/72.97932",

language = "English (US)",

volume = "2",

pages = "548--558",

journal = "IEEE Transactions on Neural Networks",

issn = "1045-9227",

publisher = "IEEE Computer Society",

number = "6",

}

TY - JOUR

T1 - Bounds on the Number of Samples Needed for Neural Learning

AU - Mehrotra, Kishan G.

AU - Mohan, Chilukuri K.

AU - Ranka, Sanjay

PY - 1991/11

Y1 - 1991/11

N2 - This paper addresses the relationship between the number of hidden layer nodes in a neural network, the complexity of a multiclass discrimination problem, and the number of samples needed for effective learning. Bounds are given for the latter. We show that Ω(min (d, n). M) boundary samples are required for successful classification of M clusters of samples using a two-hidden-layer neural network with d-dimensional inputs and n nodes in the first hidden layer.

AB - This paper addresses the relationship between the number of hidden layer nodes in a neural network, the complexity of a multiclass discrimination problem, and the number of samples needed for effective learning. Bounds are given for the latter. We show that Ω(min (d, n). M) boundary samples are required for successful classification of M clusters of samples using a two-hidden-layer neural network with d-dimensional inputs and n nodes in the first hidden layer.

UR - http://www.scopus.com/inward/record.url?scp=0026254755&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026254755&partnerID=8YFLogxK

U2 - 10.1109/72.97932

DO - 10.1109/72.97932

M3 - Article

AN - SCOPUS:0026254755

SN - 1045-9227

VL - 2

SP - 548

EP - 558

JO - IEEE Transactions on Neural Networks

JF - IEEE Transactions on Neural Networks

IS - 6

ER -

Bounds on the Number of Samples Needed for Neural Learning

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this