[Model of nerve net learning based on changes in the efficiency of excitatory and inhibitory synapses]

Based on recent data on learning neurophysiology, a dynamic model of a nervous network was created, consisting in systems of locally connected excitatory and inhibitory elements and general excitatory and inhibitory systems. Interneuronal interactions were realized by means of imitation of impulse transmission of information, spatial-temporal summation of excitatory and inhibitory influences, the effect of "disinhibition" by the mechanism, imitating the depression of inhibitory elements during overexcitation and presynaptic inhibition of inhibitory systems. When learning the model, the Hebb's principle was applied, i. e. and irreversible increase of synaptic transmission after coincidence of activation of pre- and postsynaptic neurones. The processes, imitating the elaboration, extinction and recovery of conditioned reflex were studied. Some initially unforeseen effects were revealed. The universal properties of the model are being discussed.