Forecast analysis of foreign trade based on road transportation between Türkiye and the Balkan countries

The primary objective of this study is to thoroughly analyze the forecast for road freight transport volumes between Turkey and the five Balkan countries of Bulgaria, Romania, Greece, Albania, and North Macedonia, whose trade relations are discussed, for the years 2025-2029, and to develop robust future forecasts. The dataset used in this thesis is taken from the foreign trade and road transport statistics of the Turkish Statistical Institute (TUIK). The basis of the study is a comprehensive annual dataset covering a period from January 2002 to 2024. This long-term dataset allows for the understanding of both historical trends and fluctuations. Methodologically, four different time series forecasting methods were applied in the study, and their performances were comparatively evaluated. The methods used were Support Vector Regression, Gaussian Processes, Multilayer Perceptron, and Linear Regression, with the preferred models as the primary models. The forecasting success of the models was comparatively evaluated using the error metrics Root Mean Square Error, Mean Absolute Error, and Mean Absolute Percentage Error. These assessments demonstrate that time series analysis can be a valuable tool in logistics planning and supply chain management within the Turkey-Balkans trade corridor. It is also expected to contribute to the field and offer a long-term perspective to academia and industry.