Abstract
A novel activity associated to the neurons of a SOM, called Residual Activity (RA), is defined in order to enlarge into the temporal domain the capabilities of a Self-Organizing Map for clustering and classifying the input data when it offers a temporal relationship. This novel activity is based on the biological plausible idea of partially retaining the activity of the neurons for future stages, that increases their probability to become the winning neuron for future stimuli. The proposed paper also proposes two quantifiable parameters for evaluating the performances of algorithms that aim to exploit temporal relationship of the input data for classification. Special designed benchmarks with spatio-temporal relationship are presented in which the proposed new algorithm, called TESOM (acronym for Time Enhanced SOM), has demonstrated to improve the temporal index without decreasing the quantization error.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Ruwisch, M.B.D., Purwins, H.-G.: Parallel hardware implementation of kohonen’s algorithm with an active medium. Neural Networks 6, 1147–1157 (1993)
Euliano, N.R., Principe, J.C.: Spatiotemporal self-organizing feature maps. In: Proceedings of International Joint Conference on Neural Networks, vol. 4, pp. 1900–1905 (1996)
Principe, N.E.J., Garani, S.: Principles and networks for self-organization in space-time. Neural Networks (15), 1069–1083 (2002)
Kangas, J.: Phoneme recognition using time-dependent versions of self-organizing maps. In: Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing, vol. 1, pp. 101–104 (1991)
Kohonen, T.: The hypermap architecture. Artificial Neural networks, 1357–1360 (1991)
Morrasso, P.: Self-organizing feature maps for cursive script recognition. Artificial Neural networks, 1323–1326 (1991)
Vauche, G., Mozayyani, N., Alanou, V., Dreyfus, J.F.: A spatio-temporal data-coding applied to kohonen maps. In: Proceedings of the International Conference on Artificial Neural Networks, vol. 2, pp. 75–79 (1995)
Voegtlin, T.: Recursive self-organizing maps. Neural Networks 15(8-9), 979–991 (2002)
Wiemer, J.C.: The time-organizer map algorithm: Extending the self-organizing map to spatiotemporal signals. Neural Computing 15, 1143–1171 (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Campoy, P., Vicente, C.J. (2005). Residual Activity in the Neurons Allows SOMs to Learn Temporal Order. In: Duch, W., Kacprzyk, J., Oja, E., Zadrożny, S. (eds) Artificial Neural Networks: Biological Inspirations – ICANN 2005. ICANN 2005. Lecture Notes in Computer Science, vol 3696. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11550822_59
Download citation
DOI: https://doi.org/10.1007/11550822_59
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-28752-0
Online ISBN: 978-3-540-28754-4
eBook Packages: Computer ScienceComputer Science (R0)