Contextual kohonen SOM with orthogonal weight estimator principle
We present in this paper the embedding of the Othogonal Weight Estimator (OWE) principle in Kohonen self-organizing maps (SOM). The resulting architecture is a context-independant classification system. The modification of the SOM architecture is that the weights of the SOM are computed by a MLP feds by the context of the presented pattern. We show the results on not trivial problem that underline the capacities of this new architecture.
Unable to display preview. Download preview PDF.
- 1.Fort, J.C., Pages, G.: Quantization vs Organization in the Kohonen S.O.M. In ESANN'96 Proceedings, D Facto Brussels Belgium. Bruges April 24–26, 1996.Google Scholar
- 2.Kohonen, T.: Self-Organizing Maps. Springer Series in Information Sciences, Springer, New York 1995.Google Scholar
- 3.Pican, N., Fort, J.C, and Alexandre, F.: Lateral Contribution Learning Algorithm for Multi-MLP Architecture. ESANN'94 Proceedings, D Facto Brussels Belgium. April 20–22, 1994.Google Scholar
- 4.Pican, N., Fort, J.C, and Alexandre, F.: An On-line Learning Algorithm for the Orthogonal Weight Estimation of MLP. Neural Processing Letters, D Facto Brussels Belgium. 1(1) (1994), 21–24.Google Scholar
- 5.Pican, N., Fohr, D., and Mari, J.F HMMs and OWE Neural Network for Continuous Speech Recognition. In Proceedings of International Conference on Spoken Language Processing, ICSLP. Philadelphia October 3–6, 1996.Google Scholar
- 6.Pican, N.: Intrinsic and Parallel performances of the OWE Neural Network. In Proceedings of International Conference on Artificial Neural Networks, ICANN. Bochum, Germany. 1996.Google Scholar