LoopSOM: A Robust SOM Variant Using Self-Organizing Temporal Feedback Connections

• Rafael C. Pinto UFRGS
• Paulo M. Engel UFRGS

Resumo

This paper introduces feedback connections into a previously proposed model of the simple and complex neurons of the neocortex. The original model considers only feedforward connections between a SOM (Self-Organizing Map) and a RSOM (Recurrent SOM). A variant of the SOM-RSOM pair is proposed, called LoopSOM. The RSOM sends predictions to the SOM, providing more robust pattern classification/recognition and solving ambiguities.

Referências

Berglund, E. (2009). Improved PLSOM algorithm. Applied Intelligence, pages 1–9.

Berglund, E. and Sitte, J. (2003). The parameter-less SOM algorithm. In Proc. ANZIIS, pages 159–164.

Cohen, L. (1998). An information-processing approach to infant perception and cognition. The development of sensory, motor, and cognitive capacities in early infancy, pages 277–300.

Elman, J. (1990). Finding structure in time. Cognitive science, 14(2):179–211.

Foldiak, P. and Young, M. (1995). Sparse coding in the primate cortex. The handbook of brain theory and neural networks, pages 895–898.

Hawkins, J. (2005). On Intelligence. Owl Books.

Hawkins, J. and George, D. (2006). Hierarchical Temporal Memory–Concepts, Theory, and Terminology. Whitepaper, Numenta Inc, May.

Kohonen, T. (1998). The self-organizing map. Neurocomputing, 21(1-3):1–6.

Koskela, T., Varsta, M., Heikkonen, J., and Kaski, K. (1998). Temporal sequence processing using recurrent SOM. In Knowledge-Based Intelligent Electronic Systems, 1998. Proceedings KES’98. 1998 Second International Conference on, volume 1.

Miller, J. and Lommel, P. (2006). Biomimetic sensory abstraction using hierarchical quilted self-organizing maps. In Proceedings of SPIE, volume 6384, page 63840A. SPIE.

Swarup, S., Lakkaraju, K., Klementiev, A., Ray, S., Sassen, E., and Munchen, G. (2005a). An anticipatory self-organized map for robust recognition. In Proceedings of the IASTED International conference on Artificial Intelligence and Applications (AIA’05), Innsbruck, Austria.

Swarup, S., Lakkaraju, K., Smith, N., and Ray, S. (2005b). A Recurrent Anticipatory Self-Organized Map for Robust Time Series Prediction. In The Workshop on Self-Organizing Maps (WSOM), Paris, France, Sep 5-8, 2005.

Voegtlin, T. (2002). Recursive self-organizing maps. Neural Networks, 15(8-9):979–991.