Abstract
We present a new definition of fault tolerance for an MLP by introducing the concept of a desired level of robustness. Based on this definition we propose an efficient method, called selective augmentation, which transforms a trained MLP to the fault tolerant one against a stuck_at_0 fault at the hidden neurons. We show, through an example, that the resulting networks designed by the proposed method are not only fault tolerant but also less redundant than the ones by the uniform augmentation.
Preview
Unable to display preview. Download preview PDF.
References
Feltham D. B. I., & Maly, W. (1991). Physically Realistic Fault Models for Analog CMOS Neural Networks. IEEE Journal of Solid-State Circuits, 26, 1223–1229.
Emmerson, M. D. & Damper, R. I. (1993). Determining and Improving the Fault Tolerance of Multilayer Perceptrons in a Pattern Recognition Application. IEEE Transactions on Neural Networks, 4, 788–793.
Naraghi-Pour, M., Hedge, M., & Bapat, P. (1993). Fault Tolerance Design of Feed-forward Networks. Proc. World Congress on Neural Networks(Vol. 3, pp. 568–571), Portland, USA.
Bolt, G. (1991). Fault Models for Artificial Neural Networks. Proc. International Joint Conference on Neural Networks(Vol. 2, pp. 1373–1378), Singapore.
Neti, C., Schneider, M. H., & Young, E. D. (1992). Maximally Fault Tolerant Neural Networks. IEEE Transactions on Neural Networks, 3, 14–23.
Séquin, C. H., & Clay, R. D. (1991). Fault tolerance in feedforward artificial neural networks, in P. Antognetti & V. Milutinović(eds.), Neural Networks Concepts, Applications, and Implementations(Vol. IV, pp. 111–141). Englewood Cliffs, New Jersey: Prentice Hall.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kwon, O.J., Bang, S.Y. (1997). Design of a fault tolerant multilayer perceptron with a desired level of robustness. In: Gerstner, W., Germond, A., Hasler, M., Nicoud, JD. (eds) Artificial Neural Networks — ICANN'97. ICANN 1997. Lecture Notes in Computer Science, vol 1327. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0020203
Download citation
DOI: https://doi.org/10.1007/BFb0020203
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63631-1
Online ISBN: 978-3-540-69620-9
eBook Packages: Springer Book Archive