engleski

Evaluating trend models in forecasting exports time series in Croatia

The paper discusses trend models for modelling exports and relative percentage share of exports in gross domestic product in Croatia. The article focuses on examining the adequacy of four trend models: linear, quadratic, exponential and second-order exponential model. The main emphasis is on comparison of statistical quality measures of forecasts based on the fitted models. For this purpose, forecasts quality indicators as Mean Absolute Percentage Error (MAPE), Mean Absolute Deviation (MAD), Mean Squared Deviation (MSD) and information criteria (AIC , SC, HQ) based on EViews and Minitab software output were applied.

regression trend models forecasting; Mean Absolute Percentage Error (MAPE); Mean Square Deviation (MSD); information criteria

Citirano u znanstvenim bazama: ISI, INSPEC, Journal of Economic Literatute, Current Mathematical Publications, Mathematical Review