Performance Measures for Associative Memories That Learn and Forget

Abstract

Recently, many modifications to the McCulloch/Pitts model have been proposed where both learning and forgetting occur. Given that the network never saturates (ceases to function effectively due to an overload of information), the learning updates can con(cid:173) tinue indefinitely. For these networks, we need to introduce performance measmes in addi(cid:173) tion to the information capacity to evaluate the different networks. We mathematically define quantities such as the plasticity of a network, the efficacy of an information vector, and the probability of network saturation. From these quantities we analytically compare different networks.