论文信息 - Globally-ordered topology-preserving maps achieved with a learning rule performing local weight updates only

Globally-ordered topology-preserving maps achieved with a learning rule performing local weight updates only

A new unsupervised competitive learning rule is introduced for topology-preserving map formation and vector quantization. The rule, called maximum entropy learning rule (MER), achieves a globally-ordered map by performing local weight updates only. Hence, contrary to Kohonen's self-organizing map algorithm and its many variations, no neighborhood function is needed. The rule yields an equiprobable quantization of a d-dimensional input p.d.f. Simulations are performed to show that the dynamical- and convergence properties of MER are essentially different from those of Kohonen's algorithm.

M. Hulle

[1] C. Malsburg,et al. How to label nerve cells so that they can interconnect in an ordered fashion. , 1977, Proceedings of the National Academy of Sciences of the United States of America.

[2] K. Schulten,et al. Kohonen's self-organizing maps: exploring their computational capabilities , 1988, IEEE 1988 International Conference on Neural Networks.

[3] Helge Ritter. Asymptotic level density for a class of vector quantization processes , 1991, IEEE Trans. Neural Networks.

[4] Hilbert J. Kappen,et al. Error potentials for self-organization , 1993, IEEE International Conference on Neural Networks.

[5] Teuvo Kohonen,et al. Physiological interpretationm of the self-organizing map algorithm , 1993 .

[6] Dominique Martinez,et al. On an Unsupervised Learning Rule for Scalar Quantization following the Maximum Entropy Principle , 1993, Neural Computation.

[7] Dominique Martinez,et al. On a novel unsupervised competitive learning algorithm for scalar quantization , 1994, IEEE Trans. Neural Networks.

[8] Dominique Martinez,et al. Generalized boundary adaptation rule for minimizing rth power law distortion in high resolution quantization , 1995, Neural Networks.