UDC 330.4

Authors

G.A. SHCHERBAKOV

Abstract

Modern science has a significant need for powerful mathematical methods that allow you to create and use adequate models of real processes. However, these processes in practice are formed under the uniqueness of situations determined by the non-stationary development. Traditional models, on the contrary, are based on postulates about the determinism of factors or about the possibility of obtaining experimental information about them in any volume. The article considers a progressive method of bridging such a "gap", called the "regularizing Bayesian approach." This development can not only compose a reliable mathematical platform for analyzing the situation in a changing environment, but is also able to carry out current adjustment of forecasts based on the processing of new heterogeneous and fuzzy data.

Keywords

mathematical methods, analysis, forecasting, technological development, complex dynamic systems.