Neural nets activity in design cognitive model of brain
A.-M. Truta · 2004
Neuroinformatics is a multidiscipline that results from synergetic actions of several theories such as achievement, processing, storage, transmissions recovery and diffusion of neural information. From neuroinformatics point of view, the neural complex (natural or artificial neural nets) is considered an automata with self-control, a memory machine and hemostats (hemostats represent the whole internal processes and behavior that have as a main goal - the achievement of an equilibrium state in several changes of environment). Neural nets (natural or artificial) are neural complex systems with C/sup 3/I protocol (commands, communication, control and information). Neural nets consist of cellular units strongly interconnected. Excitatory/inhibitory activities of cellular units propagate information to the entire system. Parallel information processing in these units leads to network convergence by cost function minimizing. Neural activity is described by the percentage of the excitatory/inhibitory cellular units. The excitatory activity is described as negentropy (the uncertainty parameter) and the inhibitory activity is described as posentropy (the certainty parameter).