CONTROL OF ASSOCIATIVE DYNAMICS BY MATCHING FEATURES IN CHAOTIC NEURAL-NETWORK

When patterns are stored by using associative learning in a chaotic ne ural network consisting of chaotic neurons that easily produces chaos, whether dynamic (chaotic) or static (non-chaotic) remembering occurs can be controlled by varying the parameters of the chaotic neurons. A pattern can be searched by using this dynamic remembering as a samplin g procedure. If features of the output pattern of the network become s imilar to desired features during the search, the chaotic state is cha nged to a static remembering state so that the output pattern has desi red features. In this paper, the state is controlled by using a model of presynaptic inhibition similar to that seen in real nerve cells (no t using the parameters of the chaotic neurons). Extraction and compari son of the features are performed by back-propagation networks. The ex traction of features under this method is amenable to visual represent ation of its trajectory in feature space.