Zastosowanie modelowania ekonometrycznego w prognozowaniu brakujących danych w szeregach o wysokiej częstotliwości
Application of econometrical modeling for predicting missing data in high-frequency time series

Summary/Abstract: The paper presents the results of modeling and forecasting the lack of full in-formation, with the example of electric energy demand in a certain area for half-hourly data. The forecasts were built based on models depicting three types of periodic fluctuations with the annual, weekly and 24-hourly cycles, by the means of dummy variables. Moreover, they included such variables for determining holidays and a delayed predicted variable. In the course of empiric verification of the predictions, the analysis of precision of interpolation and extrapolation forecasts was performed.

• Details
• Contents