Türkiye'deki Hafif Ticari Araç Satışlarının Makine Öğrenmesi Yöntemleriyle Tahmin Edilmesi.

Light commercial vehicles such as vans, pick-up trucks, panel vans, and minibuses are a class of vehicles used in large quantities, especially in the trade and service sectors. Changes in demand for this class of vehicles are also perceived as an indicator of the economic vitality of countries. In this study, it is thought that forecasting the sales and imports of light commercial vehicles, which is accepted as a macroeconomic indicator, will contribute to the evaluation of general economic indicators and will be useful for automotive companies operating in this market in terms of effective corporate resource planning and efficient use of resources from a micro perspective. The sales forecasting model designed in this study was created by analysing previous studies in the literature and including macroeconomic variables that are thought to affect light commercial vehicle sales in the model. The sales forecasting model designed in this study is constructed by analysing previous studies in the literature and including macroeconomic variables that are thought to affect light commercial vehicle sales in the model. Three machine learning methods, namely artificial neural network (ANN), multiple linear regression (MLR) and decision tree (DT) regression, were used to measure the forecasting success of the model. As a result of the study, the R2 value was found to be 94.6% for ANN, 64.1% for MLR, and 82.2% for DT. According to the results obtained, it is concluded that the model designed for the prediction of light commercial vehicle sales in Turkey performs very successful predictions with ANN method. [ABSTRACT FROM AUTHOR]