The McCulloch-Pitts model

Neural net theory is founded on the model of McCulloch and Pitts (1943). The article discusses the principles of the model and the associated algebra. The adaptability comes from representing the synaptic action by a variable weight which determines the degree to which a neuron should 'take notice' of firing signals that take place at the synapse concerned. The neuron is thought to take firing signals at all its synapses into account by summing their effects, both excitatory and inhibitory, and thereby determining whether it should or should not fire. The effect of a synapse is represented by a weight in the range -1 to 1. The effect on a neuron of any particular synapse is the weight if the neuron fire, 0 if not, and the product of the weight and another number if it be not known whether the neuron fires.