Finding the dependencies of the parameters of the forecast models on the characteristics of the output sets of administrative monitoring data

• Published in: 2017 IEEE 11th International Conference on Application of Information and Communication Technologies (AICT) pp. 1 - 4
• Main Authors: Lukianov, P.V., Frolov, A.I., Frolova, V.A.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-09-2017
• Subjects: administrative monitoring; automatic selection; Correlation; Data models; Forecasting; model parameters; Monitoring; organizational and technical system; predicting; Predictive models; Task analysis; Time series analysis
• DOI: 10.1109/ICAICT.2017.8686982
• EISSN: 2472-8586

Administrative monitoring is a kind of monitoring that involves the operative collection of information about the state of the management object from various sources, followed by its processing to the required level of abstraction and further provision of it as a starting point for various decision support systems and/or to the decision maker in the organizational and technical system, in the conditions of changing requirements for the structure and composition of indicators. In article questions of automated selection of model parameters in predicting the state of complex organizational and technical systems on the basis of administrative monitoring data are considered. The linear regression method is described to find the relationship between the parameters of forecasting models and the characteristics of output sets of administrative monitoring data. For some forecasting parameters, a direct dependence on the characteristics of the output data sets of administrative monitoring is determined.