Kononenko A. Structural and parametrical synthesis of prediction models for business-processes in conditions of marketing information deficiency

Subject of research: quantitative forecasting methods of economic parameters and integrated modeling of enterprise activity. Goal of research: reliability improvement of short-term forecasting based on multidimensional time series in conditions of marketing information deficiency by means of mathematical adaptive models synthesis and software tools for modelling enterprise activity. Research methods: system analysis method, probability theory, statistical analysis techniques and marketing research strategic analysis tools. Scientific results of research: Structural synthesis of exponential smoothing model method based on dynamic sequences of basic data analysis. As opposed to known ones this method allows to do suitable choice for mathematical exponential smoothing model and thereby to improve the forecast reliability for modelling results; exponential smoothing model as a result of development of adaptive parametric synthesis procedure for model’s adjustable parameters by means of its retrospective estimating. This allow to take into account personalities of the specific process to be simulated and to increase accuracy on this basis; a complicated forecasting method of enterprises activity based on strategic matrixes as a model and applicable software tools at the expense of dynamic form for research results visualisation with forecasting value of business-unit positioning at period under review. Practical importance: suggested models, methods and developed algorithms allow to improve a forecasting reliability for economic parameters of enterprise activity that in turn make possible to conduct science-based operational planning of its activity. The level for results implementation is branch. Obtained results may be used with conducting researches and in production and trade enterprises.