Variants of self-organizing maps

摘要: Self-organizing maps have a bearing on traditional vector quantization. A characteristic that makes them more closely resemble certain biological brain maps, however, is the spatial order of their responses, which formed in learning process. discussion presented basic algorithms and two innovations: dynamic weighting input signals at each cell, improves ordering when very different are used, definition neighborhoods algorithm by minimal spanning tree, provides far better faster approximation prominently structured density functions. It cautioned if used for pattern recognition decision process, it necessary to fine tune reference vectors so they directly define borders. >