Probabilistic-deterministic neutron-like structures: a base for perspective neurocomputers

A purely formalized approach is suggested which enables a designer to make neuron-like networks by calculating and without resorting to heuristics and indeterminancy. The resulting neuron-like networks possess high reliability and survivability. Even when a large proportion of constituent elements (up to a fixed threshold value) stops functioning, the network as a whole continues to work as if all the constituent elements were perfect. The main organizing principles of probabilistic-deterministic neuron-like networks (PDNN) are presented. Principles for the construction of mathematical models of PDNN functioning are also examined