Application of descriptive models to forecasting seasonal time series with gaps

In this paper were presented the results of the application of quasi-simulation methods to analysis the impact of the occurrence of systematic gaps on the accuracy of inter and extrapolative forecasts for time series with seasonal fluctuations. Forecasts were built on the basis of predictors based on descriptive models with seasonally changing parameters. Theoretical considerations will be illustrated by the empirical example. The models estimation and construction of inter- and extrapolative forecasts were done with R and Statistica 10.