Analysis of Manufacturing Industry Capacity Utilization Rate Forecasts Using Decision Tree-Based Algorithms

In the globalizing world order, different sectors are affected by global events. The manufacturing sector is one of the most critical sectors for both individual countries and the world at large. The development of this sector alleviates the economic burden on countries and helps them gain economic independence. In this context, this study aims to forecast the future values of Türkiye's manufacturing industry capacity utilization rate. To this end, monthly data from June 2013 to December 2024 were used to forecast all months of 2025. The Random Tree and Random Forest Algorithms, which are machine learning algorithms, were employed for this forecasting process. As a result of the study, the Random Tree algorithm showed a better forecasting performance with an accuracy of 83.48%. Furthermore, the Correlation Attribute Feature Selection Algorithm was used to identify the variables affecting the prediction of Türkiye's manufacturing industry capacity utilization rate. These variables were determined to be ‘Manufacture of electrical equipment’, ‘Non-Durable Consumer Goods’, ‘Investment Goods’, and ‘Durable Consumer Goods’, respectively.