Abstract
Cluster analysis applications of the SOM require it to be sensible to features, or groupings, of different sizes in the input data. On the other hand, the SOM’s behavior while the organization process is taking place also exhibits regularities of different scales, such as periodic behaviors of different frequencies, or changes of different magnitudes in the weight vectors. A method based on the discrete wavelet transform is proposed for measuring the diversity of the scales of regularities, and this diversity is compared to the performance of the SOM. We argue that if this diversity of scales is high then the algorithm is more likely to detect differently sized features of data.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Kraaijveld, M., Mao, J., Jain, A.: A non-linear projection method based on Kohonen’s topology preserving maps. In: Proceedings of the 11th Int. Conf. on Pattern Recognition, pp. 41–45 (1992)
Flexer, A.: On the Use of Self-Organizing Maps for Clustering and Visualization. Intelligent Data Analysis 5, 373–384 (2001)
Kohonen, T.: Self-organized formation of topologically correct feature maps. Biological Cybernetics 43, 59–69 (1982)
Kohonen, T.: Self-Organizing Maps. Springer, Heidelberg (1995)
Haykin, S.: Neural networks a comprehensive foundation, 2nd edn. Prentice-Hall, Englewood Cliffs (1999)
Mallat, S.: A theory for multiresolution signal decomposition: the wavelet representation. IEEE Pattern Anal. and Machine Intell. 11(7), 674–693 (1989)
Haar, A.: Zur Theorie der orthogonalen Funktionensysteme. Math. Ann. 69, 331–371 (1910)
Walnut, D.: An Introduction to Wavelet Analysis. Birkhäuser, Basel (2001)
Daubechies, I.: Ten Lectures on Wavelets Society for Industrial and Applied Mathematics (1992)
Ogden, R.: Essential Wavelets for Statistical Applications and Data Analysis. Birkhauser, Basel (1997)
Kiviluoto, K.: Topology Preservation in Self-Organizing Maps. In: Proceedings of International Conference on Neural Networks, pp. 294–299 (1996)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Mireles, V., Neme, A. (2008). Analyzing the Behavior of the SOM through Wavelet Decomposition of Time Series Generated during Its Execution. In: Kůrková, V., Neruda, R., Koutník, J. (eds) Artificial Neural Networks - ICANN 2008. ICANN 2008. Lecture Notes in Computer Science, vol 5163. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-87536-9_68
Download citation
DOI: https://doi.org/10.1007/978-3-540-87536-9_68
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-87535-2
Online ISBN: 978-3-540-87536-9
eBook Packages: Computer ScienceComputer Science (R0)