Chepynoha A. Methods of polynomial estimation of polygaussian models parameters with the moment and cumulant description.

The object of study: process of parameter estimation polygaussian models based on stochastic polynomials Kunchenko. Goal of study: the development of methods for estimation of parameters of polygaussian models with punched moment and cumulant description based on the apparatus of stochastic polynomials Kunchenko for effective implementation of distribution approximation procedures for empirical data and generation of random series. Methods of study: mathematical apparatus of probability theory and mathematical statistics, theory signal theory approximation methods, simulation (computer) modeling, mathematical and numerical methods. Methods of statistical processing obtained in the thesis, based on the use of stochastic polynomials Kunchenko system, the effectiveness of which is shown provided statistical differences in the distribution of Gaussian law. The scientific novelty of the results: first developed approximation method for empirical data by polygaussian distribution with perforated moment-cumulant description and polynomial evaluating their options, which increases the adequacy models. The practical value of the results: developed a method of synthesis of algorithms for computing the statistical parameter estimation polihausovyh models that can be used for applications of stochastic approximation and generation; Application of the developed methods for evaluating the quality of production processes; implemented test program generator of random sequences with desired properties in packages of engineering calculations. The research results implemented in manufacturing and in studying process.