Reduction of Visual Information
Kohonen network
Description: N - size of the neighborhood, N(M->0) - the neighborhood size decreases in the training process from initial value M to ZERO. G - the learning rate, G(M->0) - the learning rate decreases in training process from initial value M to ZERO. MxN - there are M*N neurons in the Competitive Layer (CL) of Kohonen network (MxN also describes the geometry of the CL).
Training patterns (Task1):
Training patterns (Task2):
Bibliography
Kohonen, T.: Self-organized formation of topologically correct feature maps. Neurocomputing: foundations of research (1982) 509-521.
Kohonen, T.: Automatic formation of topological maps of patters in a self organizing system. In: 2nd Scand. conference on Image Analysis, Oja, E., Simula, O. (eds.) (1981) 214-220.
Kangas, J., Kohonen, T., Laaksonen, J.: Variants of self-organizing maps. IEEE Transactions on Neural Networks 1 (1990) 93-99.