Since there are many artificial intelligence systems, there is a need of comparable quality assessment of their mathematical models. For this purpose, these systems can be tested on the same database source data, for which it is very convenient to use a public database of the UCI repository. This work is aimed at the study and development of model practices of the database of the UCI repository to assess the quality of mathematical models of artificial intelligence systems
The author points out the features of synergy, which using the linearity and nonlinearity and also express-es the aspects of material unity of the world in the theory, linked to the general properties of self-development of complex systems. The universe is seen as a dissipative system with periodically remov-able elements (elementary dissipative systems)
In this article we have presented the actual idea about the evolutional character in the modern science. The reason of the unity of nature is demonstrated in the particular example using cosmology and physics
On the basis of a new paradigm of applied mathematical statistics, data analysis and economic-mathematical methods are identified; we have also discussed five topical areas in which modern applied statistics is developing as well as the other statistical methods, i.e. five "growth points" – nonparametric statistics, robustness, computer-statistical methods, statistics of interval data, statistics of non-numeric data
We have analyzed the current state of the main computer-statistical methods, identified achievements and existing problems, outlined the prospects of further movement and formulated the problems to be solved. We have also discussed the Monte Carlo methods, pseudo-random numbers, simulation, bootstrap and resampling, the automated system-cognitive analysis. We have considered the applications of computer statistics in controlling and properties of statistical packages as the tools for researchers
We have allocated the basic sources of uncertainty in various industrial and economic situations. We have also considered the role and the tasks of forecasting in the management of industrial companies, particularly in the rocket and space industry. We
have discussed the methods of organizational and economic forecasting - statistical, expert, combined, including foresight and given some suggestions for improving the forecasting and planning mechanisms for practical use when creating space systems
In this article we consider Einstein's theory of gravity in relation to the Yang-Mills theory. It is shown that in Einstein's theory there exists a metric together with the Yang-Mills theory, in which the field equations are reduced to the Liouville equation describing the
evolution of stars. The mechanism of generation of stellar energy of dark energy in the processes of geometric turbulence is discussed
This article is a continuation of the works [1,2], which were devoted to the study of hydrodynamics and transport of salt ions in the experimental electrochemical cell with a rotating disk with a cation exchange membrane of exact current modes, when the condition of local electroneutrality. This article presents a mathematical model of transport of salt ions in a cell with a rotating disk with a cation exchange membrane exorbitant current regimes, taking into account electroconvection. Under these conditions, fluid dynamics depends on the ion transport process salt and described by the system of Navier-Stokes equations in cylindrical coordinate system with the electric forces
The article investigates the problem of estimation of diameters of prefractal graphs with priming stars and with full priming which contiguity of old edges in a trajectory isn't broken on spectra
We have proposed the method for testing of independence of two alternative variables on the basis of statistics of non-numeric data. The method is aimed at application in problems of statistical quality
control. Testing of independence is based on set of small samples, i.e., in the Kolmogorov’s asymptotics, when the number of unknown
parameters of the distribution increases in proportion to the data size