Abstract
An adaptation algorithm for online training is examined. For stationary tasks it can reduce the learning rate to reach the best convergence. Instead of simple annealing, it keeps the learning rate flexible, such that it can also adapt to nonstationary tasks. Different tasks, abrupt or gradual changes, and different guidance measures are discussed.
Preview
Unable to display preview. Download preview PDF.
References
Amari S. (1997), in NIPS 9, MIT Press, in press.
Opper M. (1996), Phys. Rev. Lett. 77, p. 4671–4674.
Sompolinsky H., Barkai N. & Seung H.S. (1995), in Neural Networks: The Statistical Mechanics Perspective, World Scientific, p. 105–130.
Bös S. (1995), in ICANN'95, p. 111-116, and to be submitted.
Bös S. (1996), in ICANN'96, Springer LNCS 1112, p. 89–94.
Bös S., Murata N., Amari S., & Müller K.-R. (1997), submitted.
Murata N., Bös S., Amari S., & Müller K.-R. (1997), in preparation.
Murata N., Müller K.-R., Ziehe A., & Amari S. (1997), in NIPS 9, MIT Press, in press.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Bös, S. (1997). Adaptive online learning for nonstationary problems. 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/BFb0020172
Download citation
DOI: https://doi.org/10.1007/BFb0020172
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