MEMORY PROPERTIES OF NON-FULLY INTERCONNECTED NEURAL-NETWORK

A model of a neural network consisting of two states neurons with the number K of (symmetric) synaptic connections per neuron treated as a v ariable was investigated numerically, Hebb's rule was used for storing uncorrelated patterns in the network. A maximal number of such patter ns, which can be effectively retrieved by the network and the process of deterioration of the memory, is examined as a function of the numbe r of synaptic connections per neuron. The influence of the number of n eurons in the network as well as boundary conditions for the storage c apacity of the network are discussed.