Summary

Proceedings of the 2012 International Symposium on Nonlinear Theory and its Applications

2012

Session Number:C2L-B

Session:

Number:632

Variable-sized Fast Kohonen Feature Map Associative Memory using Area Representation for Sequential Patterns—Association Ability for Sequential Patterns including Common Terms—

Junpei AMANO,  Yuko OSANA,  

pp.632-635

Publication Date:

Online ISSN:2188-5079

DOI:10.15248/proc.1.632

PDF download (388.5KB)

Summary:
In this paper, we examine the association ability for sequential patterns including common terms in the Variable-sized Fast Kohonen Feature Map Associative Memory using Area Representation for Sequential Patterns (VFKFMAM-AR-SP). In this model, the connection weight fixed and semi-fixed neurons are introduced, and the patterns that have already been learned are not destroyed and a new pattern can be memorized. Moreover, when unknown patterns are given, neurons can be added in the map layer if necessary. In this research, we examined the association ability for sequential patterns including common terms and confirmed that the association ability for sequential patterns depends on the weighting factor β in the recurrent difference vector.

References:

[1] H. Ichiki, M. Hagiwara and M. Nakagawa: “Kohonen feature maps as a supervised learning machine,” Proceedings of IEEE International Conference on Neural Networks, pp.1944-1948, 1993.

[2] N. Sakurai, M. Hattori and H. Ito: ”SOM associative memory for temporal sequences,” IEEE and INNS International Joint Conference on Neural Networks, pp.950-955, Honolulu, 2002.

[3] T. Shiratori and Y. Osana : “Kohonen feature map associative memory with area representation for sequential analog patterns,” Proceedings of IEEE and INNS International Joint Conference on Neural Networks, Hong Kong, 2008.

[4] H. Midorikawa and Y. Osana : “Fast Kohonen feature map associative memory using area representation for sequential analog patterns,” Proceedings of International Conference on Neural Information Processing, Sydney, 2010.

[5] J. Amano and Y. Osana : “Variable-sized fast Kohonen feature map associative memory using area representation for sequential patterns,” Proceedings of IEEE International Conference on System, Man and Cybernetics, Anchorage, 2011.

[6] T. Imabayashi and Y. Osana : “Variable-sized KFM associative memory with refractoriness based on area representation,” Proceedings of IEEE International Conference on System, Man and Cybernetics, San Antonio, 2009.