Learning of continuously transformed pattern cycles by an oscillatory neural network

A cognitive psychological pattern transformation in the processes of the brain is discussed. It is assumed that the transformation is processed by physiological oscillations. An oscillatory neural network is proposed to embed the continuously transformed pattern cycles in the limit cycles. The oscillation by destabilized memory points and the characteristic of a dynamic trajectory for an input pattern are shown. Continuously transformed character pattern cycles were learned efficiently by a simulation without the temporal inverse process.<
>