Temporal sequence processing using recurrent SOM
Authors:
- T. KoskelaTKT. Koskela
- M. VarstaMVM. Varsta
- J. HeikkonenJHJ. Heikkonen
- K. KaskiKKK. Kaski
- 27 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1, 290-297
Abstract
Recurrent Self-Organizing Map (RSOM) is studied in temporal sequence processing. RSOM includes a recurrent difference vector in each unit of the map, which allows storing temporal context from consec- utive input vectors fed to the map. RSOM is a mod- ification of the Temporal Kohonen Map (TKM). It is shown that RSOM learns a correct mapping from temporal sequences of a simple synthetic data, while TKM fails to learn this mapping. In addition, two case studies are presented, in which RSOM is ap- plied to EEG based epileptic activity detection and to time series prediction with local models. Results suggest that RSOM can be efficiently used in tempo- ral sequence processing.Keywords
This publication has 10 references indexed in Scilit:
- Theoretical aspects of the SOM algorithmNeurocomputing, 1998
- Neural and statistical classifiers-taxonomy and two case studiesIEEE Transactions on Neural Networks, 1997
- A recurrent self-organizing map for temporal sequence processing1997
- Self-Organizing Maps1995
- Locally recurrent globally feedforward networks: a critical review of architecturesIEEE Transactions on Neural Networks, 1994
- The temporal Kohønen mapNeural Networks, 1993
- Image compression using a multilayer neural networkPattern Recognition Letters, 1989
- A theory for multiresolution signal decomposition: the wavelet representationIEEE Transactions on Pattern Analysis and Machine Intelligence, 1989
- Self-Organization and Associative Memory1989
- Convergence properties of Kohonen's topology conserving maps: fluctuations, stability, and dimension selectionBiological Cybernetics, 1988