A New Approach to Kanerva ' s Sparse Distributed

101 A New Approach to Kanerva's Sparse Distributed Memory Tim A. Hely, David J. Willshaw, and Gillian M. Hayes Abstract|The Sparse Distributed Memory (SDM)[1] was originally developed to tackle the problem of storing large binary data patterns. The model succeeded well in storing random input data. However its e ciency, particularly in handling non{random data, was poor. In its original form it is a static and in exible system. Most of the recent work on the SDM has concentrated on improving the e ciency of a modi ed form of the SDM introduced by Prager [2], which treats the memory as a single-layer neural network.(See also [3], [4], [5], [6].) This paper introduces an alternative SDM, the SDM Signal model [7] which retains the essential characteristics of the original SDM, whilst providing the memory with a greater scope for plasticity and self-evolution. By removing many of the problematic features of the original SDM the new model is not as dependent upon a priori input values. This gives it an increased robustness to learn either random or correlated input patterns. The improvements in this new SDM signal model should be also of bene t to modi ed SDM neural network models. Keywords|SDM,Kanerva,memory,model,signal.