Maximally Fault Tolerant Neural Networks

Chalapathy Neti, Michael H. Schneider, Eric D. Young

Research output: Contribution to journalArticle

Abstract

An application of neural network modeling is described for generating hypotheses about the relationships between response properties of neurons and information processing in the auditory system. Specifically, the goal is to study response properties that are useful for extracting sound localization information from directionally selective spectral filtering provided by the pinna. For studying sound localization based on spectral cues provided by the pinna, a feedforward neural network model with a guaranteed level of fault tolerance is introduced. The notions of fault tolerance and uniform fault tolerance in a neural network are formally defined and a method is described to ensure that the estimated network exhibits fault tolerance. The problem of estimating weights for such a network is formulated as a largescale nonlinear optimization problem. Numerical experiments indicate that solutions with uniform fault tolerance exist for the pattern recognition problem considered in this paper. Furthermore, solutions derived by introducing fault tolerance constraints have better generalization properties than solutions obtained via nnconstrained back propagation.

Original languageEnglish (US)
Pages (from-to)14-23
Number of pages10
JournalIEEE Transactions on Neural Networks
Volume3
Issue number1
DOIs
StatePublished - 1992

Fingerprint

Fault tolerance
Fault Tolerance
Fault-tolerant
Neural Networks
Neural networks
Acoustic waves
Network Modeling
Feedforward neural networks
Feedforward Neural Networks
Back Propagation
Nonlinear Optimization
Backpropagation
Information Processing
Neural Network Model
Neurons
Pattern Recognition
Pattern recognition
Nonlinear Problem
Neuron
Filtering

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Science Applications
  • Software
  • Computational Theory and Mathematics
  • Hardware and Architecture
  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Theoretical Computer Science

Cite this

Maximally Fault Tolerant Neural Networks. / Neti, Chalapathy; Schneider, Michael H.; Young, Eric D.

In: IEEE Transactions on Neural Networks, Vol. 3, No. 1, 1992, p. 14-23.

Research output: Contribution to journalArticle

Neti, C, Schneider, MH & Young, ED 1992, 'Maximally Fault Tolerant Neural Networks', IEEE Transactions on Neural Networks, vol. 3, no. 1, pp. 14-23. https://doi.org/10.1109/72.105414
Neti, Chalapathy ; Schneider, Michael H. ; Young, Eric D. / Maximally Fault Tolerant Neural Networks. In: IEEE Transactions on Neural Networks. 1992 ; Vol. 3, No. 1. pp. 14-23.
@article{c512ba988bbb4441bd26dce8110ae771,
title = "Maximally Fault Tolerant Neural Networks",
abstract = "An application of neural network modeling is described for generating hypotheses about the relationships between response properties of neurons and information processing in the auditory system. Specifically, the goal is to study response properties that are useful for extracting sound localization information from directionally selective spectral filtering provided by the pinna. For studying sound localization based on spectral cues provided by the pinna, a feedforward neural network model with a guaranteed level of fault tolerance is introduced. The notions of fault tolerance and uniform fault tolerance in a neural network are formally defined and a method is described to ensure that the estimated network exhibits fault tolerance. The problem of estimating weights for such a network is formulated as a largescale nonlinear optimization problem. Numerical experiments indicate that solutions with uniform fault tolerance exist for the pattern recognition problem considered in this paper. Furthermore, solutions derived by introducing fault tolerance constraints have better generalization properties than solutions obtained via nnconstrained back propagation.",
author = "Chalapathy Neti and Schneider, {Michael H.} and Young, {Eric D.}",
year = "1992",
doi = "10.1109/72.105414",
language = "English (US)",
volume = "3",
pages = "14--23",
journal = "IEEE Transactions on Neural Networks and Learning Systems",
issn = "2162-237X",
publisher = "IEEE Computational Intelligence Society",
number = "1",

}

TY - JOUR

T1 - Maximally Fault Tolerant Neural Networks

AU - Neti, Chalapathy

AU - Schneider, Michael H.

AU - Young, Eric D.

PY - 1992

Y1 - 1992

N2 - An application of neural network modeling is described for generating hypotheses about the relationships between response properties of neurons and information processing in the auditory system. Specifically, the goal is to study response properties that are useful for extracting sound localization information from directionally selective spectral filtering provided by the pinna. For studying sound localization based on spectral cues provided by the pinna, a feedforward neural network model with a guaranteed level of fault tolerance is introduced. The notions of fault tolerance and uniform fault tolerance in a neural network are formally defined and a method is described to ensure that the estimated network exhibits fault tolerance. The problem of estimating weights for such a network is formulated as a largescale nonlinear optimization problem. Numerical experiments indicate that solutions with uniform fault tolerance exist for the pattern recognition problem considered in this paper. Furthermore, solutions derived by introducing fault tolerance constraints have better generalization properties than solutions obtained via nnconstrained back propagation.

AB - An application of neural network modeling is described for generating hypotheses about the relationships between response properties of neurons and information processing in the auditory system. Specifically, the goal is to study response properties that are useful for extracting sound localization information from directionally selective spectral filtering provided by the pinna. For studying sound localization based on spectral cues provided by the pinna, a feedforward neural network model with a guaranteed level of fault tolerance is introduced. The notions of fault tolerance and uniform fault tolerance in a neural network are formally defined and a method is described to ensure that the estimated network exhibits fault tolerance. The problem of estimating weights for such a network is formulated as a largescale nonlinear optimization problem. Numerical experiments indicate that solutions with uniform fault tolerance exist for the pattern recognition problem considered in this paper. Furthermore, solutions derived by introducing fault tolerance constraints have better generalization properties than solutions obtained via nnconstrained back propagation.

UR - http://www.scopus.com/inward/record.url?scp=0026679384&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026679384&partnerID=8YFLogxK

U2 - 10.1109/72.105414

DO - 10.1109/72.105414

M3 - Article

C2 - 18276402

AN - SCOPUS:0026679384

VL - 3

SP - 14

EP - 23

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

SN - 2162-237X

IS - 1

ER -