Forecasting characteristics of time series to support managerial decision making process in production-and-economic systems

Studies the issue of forecasting the parameter values that describe production-and-economic systems and the projects that are realized in such systems; it examines the preparation of statistical data with gaps and suggests the advantage of using different time slots of the collected data for their further joint use in prognostic models. The article reviews the applicability of the existing forecasting methods in production-and-economic systems, demonstrates the application of regression analysis method and support vector machine. The experiment shows that the accuracy of gained forecasts can be taken into account in risk metrics. Since the additional information about the nature of data improves forecasting process it is recommended to use innovation and S-curve methods and semi-supervised models.