Information flow
Neural network computers are intended to be more closely analogous to the operation of the human brain than conventional serial computers. The function of neurones is simulated by components which are arranged in a network: several inputs represent synaptic connections which cause the neurone to learn to fire when a particular threshold signal level is reached.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Experimental neural network computer: Miles Dobson, 1991. This model was constructed so that the logical structure, the switching of the network, and the adjustment of the threshold levels could all be seen and understood.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.