Chaotic Models of the Hippocampus for Dynamic Pattern Recognition

The paper carried out an analysis of using systems with chaotic dynamics to solve the problem of dynamic pattern recognition. We reviewed the existing chaotic models of the hippocampus for storage, coding and retrieval of dynamic information. An episodic chaotic associative memory models proposed by Y. Osana, a hippocampus-neocortex model proposed by T. Kuremoto, and Tsuda's hippocampus model are considered in detail. The first two of these models incorporate Aihara’s chaotic neural networks. We compared the selected models based on the simulation results. It is shown that chaotic dynamics is necessary in order to take into account the context of dynamic pattern recognition problems. Trends of further modification of the models are also proposed.