REAL-TIME SYNTHESIS OF SPARSELY INTERCONNECTED NEURAL ASSOCIATIVE MEMORIES

The problem of implementing associative memories using sparsely interc onnected generalized Brain-State-in-a-Box (gBSB) network is addressed in this paper. In particular, a ''designer'' neural network that synth esizes the associative memories is proposed. An upper bound on the tim e required for the designer network to reach a solution is determined. A neighborhood criterion with toroidal geometry for the cellular gBSB network is analyzed, in which the number of adjacent cells is indepen dent of the generic cell location. A design method of neural associati ve memories with prespecified interconnecting weights is presented. Th e effectiveness of the proposed synthesis method is demonstrated with numerical examples. (C) 1998 Elsevier Science Ltd. All rights reserved .