Forecasting non-residents' monthly entries to Tunisia and accuracy comparison of time-series methods
Forecasting non-residents' monthly entries to Tunisia and accuracy comparison of time-series methods
The main objective of this article is to model and forecast non-residents' entries to the Tunisian tourist destination (North Africa). Five methods are applied, namely seasonal naïve 1, seasonal naïve 2, Holt–Winters seasonal multiplicative model, SARIMA and SARIMAX, incorporating the effects of political shocks Tunisia went through during the last 10 years. Although no clear-cut conclusions can be drawn and despite seasonal naïve 1 and Holt–Winters satisfactory performances, in-sample and out-of-sample forecast error calculations reveal that the SARIMAX is overall the most accurate method. On this basis, optimistic and pessimistic forecast scenarios are settled and managerial recommendations are proposed with the objective of improving Tunisian tourism policies.
CITATION: Klabi, Fethi. Forecasting non-residents' monthly entries to Tunisia and accuracy comparison of time-series methods . : Taylor & Francis Group , 2014. Journal of North African Studies,Vol. 19, No. 5, December 2014, pp. 770-791 - Available at: https://library.au.int/forecasting-non-residents-monthly-entries-tunisia-and-accuracy-comparison-time-series-methods-36