Abstract
A modified backpropagation algorithm that minimizes the sensitivity to weight errors is presented. Multi-Layer Perceptrons (MLPs) trained with this algorithm are more tolerant to weight deviations compared to those obtained with a classical algorithm while the other performance figures are similar. Thus the algorithm is useful for MLPs that are going to be mapped on a physical implementation that can be affected by weight imprecision.
Preview
Unable to display preview. Download preview PDF.
References
M. Stevenson, R. Winter, B. Widrow, “Sensitivity of Neural Networks to Weight Errors”, IEEE Trans. on Neural Networks, vol.1, no.1, pp. 71–80, Mar 1990.
C. Alippi, V. Piuri, M. Sami, “Sensitivity to Errors in Artificial Neural Networks: a Behavioral Approach”, in Proc. IEEE Int. Symp. on Circuits & Systems, pp. 459–462, May 1994.
C. Chiu K Mehrotra, C. K. Mohan, S. Ranka, “Modifying Training Algorithms for Improved Fault Tolerance”, in Proc. IEEE Int. Conf. on Neural Networks, vol.1 pp. 333–338, Jun 1994.
C. Lin, I. Wu, “Maximizing Fault Tolerance in Multilayer Neural Networks”, in Proc. IEEE Int. Conf. on Neural Networks, pp. 419–424, Jun 1994.
H. Elsimary, S. Mashali, A. Darwish, S. Shasheen, ”Performance Evaluation of a Novel Fault Tolerance Training Algorithm”, in Proc. Int. Conf. on Electronics, Circuits & Systems, pp.566–570, Dec. 1994.
J.Y. Choi, C. Choi, ”Sensitivity Analisis of Multilayer Perceptron with Differentiable Activation Functions”, IEEE Trans. on Neural Networks, vol.3, no.1, pp.101–107, Jan 1992.
R.P. Lippmann, ”An Introduction to Computing with Neural Nets”, IEEE ASSP Magazine, pp. 4–22, Apr 1987.
A. Grace, ”Optimization Toolbox”, in Matlab User's Guide. The MathWorks Inc, Jan 1994.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Bernier, J.L., Ortega, J., Prieto, A. (1997). A modified backpropagation algorithm to tolerate weight errors. In: Mira, J., Moreno-Díaz, R., Cabestany, J. (eds) Biological and Artificial Computation: From Neuroscience to Technology. IWANN 1997. Lecture Notes in Computer Science, vol 1240. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0032535
Download citation
DOI: https://doi.org/10.1007/BFb0032535
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63047-0
Online ISBN: 978-3-540-69074-0
eBook Packages: Springer Book Archive