Two-stage parallel partial retraining scheme for defective multi-layer neural networks

K. Yamamori, T. Abe, S. Horiguchi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

We address a high-speed defect compensation method for multi-layer neural networks implemented in hardware devices. To compensate stuck defects of the neurons and weights, we have proposed a partial retraining scheme that adjusts the weights of a neuron affected by stuck defects between two layers by a backpropagation (BP) algorithm. Since the functions of defect compensation can be achieved by using learning circuits, we can save chip area. To reduce the number of weights to adjust, it also leads to high-speed defect compensation. We propose a two-stage partial retraining scheme to compensate input unit stuck defects. Our simulation results show that the two-stage partial retraining scheme can be about 100 times faster than whole network retraining by the BP algorithm.

Original languageEnglish
Title of host publicationProceedings - 4th International Conference/Exhibition on High Performance Computing in the Asia-Pacific Region, HPC-Asia 2000
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages642-647
Number of pages6
ISBN (Electronic)0769505902, 9780769505909
DOIs
Publication statusPublished - 2000 Jan 1
Externally publishedYes
Event4th International Conference/Exhibition on High Performance Computing in the Asia-Pacific Region, HPC-Asia 2000 - Beijing, China
Duration: 2000 May 142000 May 17

Publication series

NameProceedings - 4th International Conference/Exhibition on High Performance Computing in the Asia-Pacific Region, HPC-Asia 2000
Volume2

Other

Other4th International Conference/Exhibition on High Performance Computing in the Asia-Pacific Region, HPC-Asia 2000
CountryChina
CityBeijing
Period00/5/1400/5/17

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Networks and Communications
  • Computer Science Applications

Fingerprint Dive into the research topics of 'Two-stage parallel partial retraining scheme for defective multi-layer neural networks'. Together they form a unique fingerprint.

Cite this