Kashcheyeva V. Information technology of analyzing multicomponent processes by time rows based on interval predictive models

Object of research is analysis process of multicomponent processes by time rows; subject of research is interval predictive models for information technologies of analyzing multicomponent processes by time rows; objective is to increase accuracy of prediction by developing new and improving present models and methods, on their basis create information technologies of analyzing multicomponent processes by time rows; methods are system analysis, control theory, mathematical statistics, interval analysis, correlation and regression analysis, retrospective evaluation; novelty is that for the first time it was developed combined predictive model of multicomponent process in conditions of limited data sampling, parametric synthesis of which is based on exchange of internal parameters of Theil-Wage model and Singular spectrum analysis; improved is information technology of multicomponent process analysis by expanding simulation tools on account of interval recurrent models and method of interval approximation of dynamic characteristics of time row fragments; developed was the model of exponential smoothing of interval statistic data by means of proposed structural parametric synthesis; results are the following: developed models and methods allowed increasing prediction, in particular, when using combined model prediction accuracy has increased by a factor of three.