Abstract
Technological breakthroughs have transformed communication and taken transportation, health, and commerce to an unprecedented level. In this way, sudden developments have rapidly affected all countries. In this context, analysis methods are changing compared to the past, and annual analyses fail to catch the trend even for macroeconomic indicators. In this paper, new artificial intelligence-based estimation methods were used to see the future trend of export volume, and their estimation performances were compared by adding them to the classical econometric method. Historical quarterly data from 2013 to 2021 were used in the training and testing phases of the models. For this purpose, the variables of gross domestic product, foreign direct investment, and dollar exchange rate, which affect the export volume, were determined as inputs in estimating the export volume. According to the analysis results, support vector machine model for predicting export volume in Turkey. This study can provide an essential basis for policymakers to export estimation and formulate their export-enhancing policies effectively.
Publisher
Bitlis Eren Universitesi Fen Bilimleri Dergisi
Reference44 articles.
1. [1] Worldbank, “Global Economic Prospects,” Washington, DC, 2022. doi: 10.1596/978-1-4648-1758-8.
2. [2] TURKSTAT, “Yıllık Gayrisafi Yurt İçi Hasıla, 2020,” 2021. https://data.tuik.gov.tr/Bulten/Index?p=Yillik-Gayrisafi-Yurt-Ici-Hasila-2020-37184#:~:text=Yıllık verilere dayalı olarak hesaplanan,milyar 883 milyon TL oldu.
3. [3] Reuters, “Turkey’s economy grew 11% in 2021; to cool to 3.5% in 2022,” 2022. https://www.reuters.com/markets/asia/turkeys-economy-grew-11-2021-cool-35-2022-2022-02-22/
4. [4] D. Guleryuz, “Forecasting outbreak of COVID-19 in Turkey; Comparison of Box–Jenkins, Brown’s exponential smoothing and long short-term memory models,” Process Saf. Environ. Prot., vol. 149, pp. 927–935, 2021, doi: https://doi.org/10.1016/j.psep.2021.03.032.
5. [5] B. Efe, M. Kurt, and Ö. F. Efe, “Hazard analysis using a Bayesian network and linear programming,” Int. J. Occup. Saf. Ergon., vol. 26, no. 3, pp. 573–588, Jul. 2020, doi: 10.1080/10803548.2018.1505805.