Shatovs'ka T. Mathematical models for determining an efficiency of the biased estimation methods

Dissertation is devoted to formalized selection of effective biased estimation method of model's parameters for kvazistasionarity process. Mathematical models for detemining an accuracy of biased estimation methods, which depends from level of multicollinearity, variance of handicapes and entrance variable is developed. There were developed a generalized criterion of choice the most accuracy method, a new imitating model for research the efficiency of biased estimation method, a method of sequential ridge regression. The results in the theory of statistical estimation, mathematical modeling were received. The results are used by development of model for degree's estimation of river's pollution and model of an influence estimation of a humid-ity level of seeds to the initial growth of shoots. Spheres of use are the technical, biotechnical and ecological systems.