A faster learning neural network classifier using selective backpropagation

The problem of saturation in neural network classification problems is discussed. The listprop algorithm is presented which reduces saturation and dramatically increases the rate of convergence. The technique uses selective application of the backpropagation algorithm, such that training is only...

Full description

Bibliographic Details
Main Author: Craven, Michael P.
Format: Conference or Workshop Item
Published: 1997
Subjects:
Online Access:https://eprints.nottingham.ac.uk/1901/
_version_ 1848790682071728128
author Craven, Michael P.
author_facet Craven, Michael P.
author_sort Craven, Michael P.
building Nottingham Research Data Repository
collection Online Access
description The problem of saturation in neural network classification problems is discussed. The listprop algorithm is presented which reduces saturation and dramatically increases the rate of convergence. The technique uses selective application of the backpropagation algorithm, such that training is only carried out for patterns which have not yet been learnt to a desired output activation tolerance. Furthermore, in the output layer, training is only carried out for weights connected to those output neurons in the output vector which are still in error, which further reduces neuron saturation and learning time. Results are presented for a 196-100-46 Multi-Layer Perceptron (MLP) neural network used for text-to-speech conversion, which show that convergence is achieved for up to 99.7% of the training set compared to at best 94.8% for standard backpropagation. Convergence is achieved in 38% of the time taken by the standard algorithm.
first_indexed 2025-11-14T18:16:29Z
format Conference or Workshop Item
id nottingham-1901
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T18:16:29Z
publishDate 1997
recordtype eprints
repository_type Digital Repository
spelling nottingham-19012020-05-04T20:33:25Z https://eprints.nottingham.ac.uk/1901/ A faster learning neural network classifier using selective backpropagation Craven, Michael P. The problem of saturation in neural network classification problems is discussed. The listprop algorithm is presented which reduces saturation and dramatically increases the rate of convergence. The technique uses selective application of the backpropagation algorithm, such that training is only carried out for patterns which have not yet been learnt to a desired output activation tolerance. Furthermore, in the output layer, training is only carried out for weights connected to those output neurons in the output vector which are still in error, which further reduces neuron saturation and learning time. Results are presented for a 196-100-46 Multi-Layer Perceptron (MLP) neural network used for text-to-speech conversion, which show that convergence is achieved for up to 99.7% of the training set compared to at best 94.8% for standard backpropagation. Convergence is achieved in 38% of the time taken by the standard algorithm. 1997 Conference or Workshop Item PeerReviewed Craven, Michael P. (1997) A faster learning neural network classifier using selective backpropagation. In: Fourth IEEE International Conference on Electronics, Circuits and Systems, 15-18 December 1997, Cairo, Egypt. neural networks ANN satuation convergence backpropagation backprop text-to-speech classification classifier
spellingShingle neural networks
ANN
satuation
convergence
backpropagation
backprop
text-to-speech
classification
classifier
Craven, Michael P.
A faster learning neural network classifier using selective backpropagation
title A faster learning neural network classifier using selective backpropagation
title_full A faster learning neural network classifier using selective backpropagation
title_fullStr A faster learning neural network classifier using selective backpropagation
title_full_unstemmed A faster learning neural network classifier using selective backpropagation
title_short A faster learning neural network classifier using selective backpropagation
title_sort faster learning neural network classifier using selective backpropagation
topic neural networks
ANN
satuation
convergence
backpropagation
backprop
text-to-speech
classification
classifier
url https://eprints.nottingham.ac.uk/1901/