What Generalizations of the Self-Organizing Map Make Sense?

The number of researchers working on the Self-Organizing Map (SOM) for the present is at least on the order of 1500, and many variants of the basic. model have already been suggested. This presentation tends to clarify the essence of the SOM, to set up its theory in the most fundamental form, and to point out what the computing functions are or should be (e.g., various structures of the network, variants of the cell function, acceleration of learning etc.). After that it will be easier to see along what lines the modifications should be developed. In particular, generalization of the static SOM into a self-organizing array of dynamic operators for sequential data is discussed.

[1]  Jouko Lampinen,et al.  Self-Organizing Maps for Spatial and Temporal AR Models , 1989 .

[2]  J. Makhoul,et al.  Vector quantization in speech coding , 1985, Proceedings of the IEEE.

[3]  Teuvo Kohonen,et al.  Things you haven't heard about the self-organizing map , 1993, IEEE International Conference on Neural Networks.

[4]  Teuvo Kohonen,et al.  Generalizations of the self-organizing map , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).

[5]  Teuvo Kohonen,et al.  Self-organization and associative memory: 3rd edition , 1989 .

[6]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[7]  Jorma Laaksonen,et al.  Variants of self-organizing maps , 1990, International 1989 Joint Conference on Neural Networks.

[8]  T. Kohonen.  Self-organized formation of topographically correct feature maps , 1982 .

[9]  Teuvo Kohonen.  THE HYPERMAP ARCHITECTURE , 1991 .

[10]  András Lörincz,et al.  Topology Learning Solved by Extended Objects: A Neural Network Model , 1994, Neural Computation.