Abstract
The Kohonen Algorithm was extended by (a) a coupling between node and input space based upon gravitational principles and (b) a controling mechanism based upon two-point correlation functions. The extended algorithm avoids optimization of the learning rate and neighbourhood parameters. Applications are given for both a 3-dimensional example data set with mixed topologies and a 13-dimensional data set of a particle physics experiment.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Kohonen, T.: Self-Organization and Associative Memory. Springer Series in Information Sciences 8, 1984, pp. 125
Fritzke, B.: Growing Cell Structures — a self-organizing network for unsupervised and supervised learning. Neural Networks 7 (1994) 1441–1460
Martinetz, T., Schulten, K.: A „neural-gas” network learns topologies. Artificial Neural Networks, Editors: Kohonen, T., Mäkisara, K., Simula, O., Kangas, J., North-Holland, Amsterdam, 1991, pp. 397
Behrmann, K.-O.: Leistungsuntersuchung des “Dynamischen Link-Matchings” und Vergleich mit dem Kohonen-Algorithmus. Diploma Thesis, Institut für Neuroinformatik, Ruhr-Universität Bochum, 1993
Lange, J. S.: Extraktion von Bremsstrahlungsereignissen in Proton-Proton-Reaktionen durch Anwendung künstlicher neuronaler Netze. Ph.D. Thesis, Institut für Experimentalphysik I, Ruhr-Universität Bochum, 1994
Lange, J. S. et al.: Cluster-Gravitation — An Extension to the Kohonen Map for the Identification of pp-Bremsstrahlung at COSY. 4th Int. Workshop on Software Engineering, Artificial Intelligence and Expert Systems for High Energy and Nuclear Physics AIHENP 95, Pisa, Apr 4–8, 1995, (in press)
Einstein, A.: Die Grundlage der allgemeinen Relativitätstheorie. Ann. d. Phys.49 (1916) 769
The COSY-TOF-Collaboration: Physica Scripta 48 (1993) 226
Hermanowski, P.: Ph. D. Thesis, Institut für Experimentalphysik I, Ruhr-Universität Bochum, 1996, (in preparation)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Lange, J.S., Freiesleben, H. (1996). A parameter-free non-growing self-organizing map based upon gravitational principles: Algorithm and applications. In: von der Malsburg, C., von Seelen, W., Vorbrüggen, J.C., Sendhoff, B. (eds) Artificial Neural Networks — ICANN 96. ICANN 1996. Lecture Notes in Computer Science, vol 1112. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61510-5_139
Download citation
DOI: https://doi.org/10.1007/3-540-61510-5_139
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61510-1
Online ISBN: 978-3-540-68684-2
eBook Packages: Springer Book Archive