V. V. Kolushov - Ph.D. (Eng.), Associate Professor, Department of «Нigher mathematics», Ufa State Aviation Technical University (Ufa). E-mail: KVV@ufanet.ru
A. V. Savelyev - Senior Research Scientist, Editor of the Journal «Neurocomputers: development, application», patent agency «©Uniquely honest patenting», www.patenttt.narod.ru. E-mail: gmkristo@rambler.ru

In the paper the substantiation of the possibility and necessity of another order of information content in the recorded empirical data was making [1]. This allows us apart from the traditional Data mining to produce not only the direct receipt of the findings from the data, extract knowledge, but along with it and produce their active construction of the initial data, one way or another, contained in the empirical data. Propose another alternative way to build (construction) of non-analytic systems using expert data received by full-scale facility as the primary and the continuation of their heuristic with obtaining a numerical and graphical-shaped new data with the prediction of the systems behavior [2]. Necessary supercomputer implementation could be avoided by the use of previously developed information СISK-technologies of «viewport». [2] Also the «floating» algorithm for computing the maximum significant of impact intrastructural-dotted interactions of object partitions [1-2] was suggested by the authors contribute to the reduction in computation time. It allows to replace the solution of the inverse problem of calculating the spatial localization of sources on the parameters of the field distribution by the solution is much simpler problem. This is the simulation problem of the external field by reducing the dimension of the problem and then «fit» of the output parameters of the model results to the existing distribution of field characteristics of the electromagnetic field around the object. Bust is made on the basis of cut-off least influence points and the functional «consolidation» of the most affecting points. The description of the work developed by the authors of algorithms and programs for 4D-visualization and animation of axons is resulted. Algorithms that are based on these models allow us to represent the spike as a collective object of cumulative interacting microcomponents from different sites (channel groups domain) of neural membrane [3]. Such a structure is embodied in the model of the neural network processor [3], reproducing complex spike being by superposition of discrete interacting microcomponents from various local domain of nano-group of neuronal membrane channels. Such spikes is smoothed and dynamically changes the degree of smoothness as it propagates. In the fig. 2 the local smoothing spike components and reconfiguration of the excited region as it moves along the axon are shown that is not reflected in the analytical models of nerve impulse propagation were described by systems of differential equations. In contrast to the known soliton models, it is seen that the initial spread of the spike is extended, and not monotonous narrowing, at the same time, increases its intensity, which more accurately reflects reality. This makes it possible not only to improve the accuracy of simulation, but also to retrieve new data, namely, to observe previously unknown effects. The methodology developed by the authors, specific to a super-constructed simulated living facility, may be general enough to apply to any tissue [4]. This methodology is a conceptual and can be used as a general approach for modeling highly complex systems of space-time and functional dynamics, including morphological, as well as the expansion of the concept of paradigm neurocomputer.