TOPOLOGY LEARNING SOLVED BY EXTENDED OBJECTS - A NEURAL-NETWORK MODEL

It is shown that local, extended objects of a metrical topological spa ce shape the receptive fields of competitive neurons to local filters. Self-organized topology learning is then solved with the help of Hebb ian learning together with extended objects that provide unique inform ation about neighborhood relations. A topographical map is deduced and is used to speed up further adaptation in a changing environment with the help of Kohonen-type learning that teaches the neighbors of winni ng neurons as well.