PATTERN ASSOCIATION AND RETRIEVAL IN A CONTINUOUS NEURAL SYSTEM

This paper studies the behavior of a large body of neurons in the cont inuum limit. A mathematical characterization of such systems is obtain ed by approximating the inverse input-output nonlinearity of a cell (o r an assembly of cells) by three adjustable linearized sections. The a ssociative spatio-temporal patterns for storage in the neural system a re obtained by using approaches analogous to solving space-time field equations in physics. A noise-reducing equation is also derived from t his neural model. In addition, conditions that make a noisy pattern re trievable are identified. Based on these analyses, a visual cortex mod el is proposed and an exact characterization of the patterns that are storable in this cortex is obtained. Furthermore, we show that this mo del achieves pattern association that is invariant to scaling, transla tion rotation and mirror-reflection.