Development of Combined Information Technology for Time Series Prediction

Abstract

The task of designing information technology for time series forecasting, that bases on fuzzy expert evaluations was considered. A forecasting model, part of which is an expert’s unit, were proposed. The algorithm of synthesis predictive scheme based on the basic predictive models was developed. To determine expert evaluation of the forecast value, the task of forecasting was seen as the problem of numerical evaluation of object. The rules for determining the collective numerical evaluations, that are based on fuzzy expert assessments were developed. The developed rules take into account coefficients of experts’ competence and also their degree of confidence for their own assessments. The approaches to determining the competence coefficients members of the expert group were systematized. The analysis of features for designing information-analytical system of time series forecasting were done. The structural diagram of the analytical block of information-analytical system for time series prediction, that based on the fuzzy expert estimates, was itemized. The designed information technology should be used for time series forecasting in cases where it is necessary to take account the impact, on the process that is studied, of temporary, informal factors.