IEICE TRANSACTIONS on Information
A Learning Algorithm for Fault Tolerant Feedforward Neural Networks
Summary :
A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.
- Publication
- IEICE TRANSACTIONS on Information Vol.E80-D No.1 pp.21-27
- Publication Date
- 1997/01/25
- Publicized
- Online ISSN
- DOI
- Type of Manuscript
- Special Section PAPER (Special Issue on Fault-Tolerant Computing)
- Category
- Redundancy Techniques
Authors
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Nait Charif HAMMADI, Hideo ITO, "A Learning Algorithm for Fault Tolerant Feedforward Neural Networks" in IEICE TRANSACTIONS on Information,
vol. E80-D, no. 1, pp. 21-27, January 1997, doi: .
Abstract: A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.
URL: https://global.ieice.org/en_transactions/information/10.1587/e80-d_1_21/_p
Copy
@ARTICLE{e80-d_1_21,
author={Nait Charif HAMMADI, Hideo ITO, },
journal={IEICE TRANSACTIONS on Information},
title={A Learning Algorithm for Fault Tolerant Feedforward Neural Networks},
year={1997},
volume={E80-D},
number={1},
pages={21-27},
abstract={A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.},
keywords={},
doi={},
ISSN={},
month={January},}
Copy
TY - JOUR
TI - A Learning Algorithm for Fault Tolerant Feedforward Neural Networks
T2 - IEICE TRANSACTIONS on Information
SP - 21
EP - 27
AU - Nait Charif HAMMADI
AU - Hideo ITO
PY - 1997
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E80-D
IS - 1
JA - IEICE TRANSACTIONS on Information
Y1 - January 1997
AB - A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.
ER -
English
