Temporal plasticity in self-organizing networks

We propose a principle that adds temporal plasticity to self-organizing networks. The algorithm uses activity diffusion to couple space and time into a single set of dynamics that can help disambiguate the static spatial information with temporal information. The approach has been successfully applied to the neural gas algorithm. We present a simple temporal example which illustrates the fundamentals of the network as well as comparing the results of our approach vs. the neural gas algorithm as applied to time-series prediction of a chaotic signal