AN ARTIFICIAL NEURAL-NETWORK SYSTEM FOR TEMPORAL-SPATIAL SEQUENCE PROCESSING

An artificial network is described that can learn, recognize, and gene rate higher-order temporal-spatial sequences. It consists of three par ts: (1) comparator units, (2) a parallel array of artificial neural ne tworks that are derived from the visual-vestibular networks of the sna il Hermissenda, as well as hippocampal neuroanatomy, and (3) delayed f eedback lines from the output of the system to the neural network laye r. Its advantages include short training time, fast and accurate retri evals, toleration of spatial noise and temporal gaps in test sequences , and ability to store a large number of temporal sequences consisting of non-orthogonal spatial patterns.