Sensitivity of feedforward neural networks to weight errors
- 1 March 1990
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks
- Vol. 1 (1) , 71-80
- https://doi.org/10.1109/72.80206
Abstract
An analysis is made of the sensitivity of feedforward layered networks of Adaline elements (threshold logic units) to weight errors. An approximation is derived which expresses the probability of error for an output neuron of a large network (a network with many neurons per layer) as a function of the percentage change in the weights. As would be expected, the probability of error increases with the number of layers in the network and with the percentage change in the weights. The probability of error is essentially independent of the number of weights per neuron and of the number of neurons per layer, as long as these numbers are large (on the order of 100 or more).Keywords
This publication has 3 references indexed in Scilit:
- Neural nets for adaptive filtering and adaptive pattern recognitionComputer, 1988
- MADALINE RULE II: a training algorithm for neural networksPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1988
- Parallel Distributed ProcessingPublished by MIT Press ,1986