DOI: 10.52693/jsas.1926730 ISSN: 2718-0999

Examining the Performance of Machine Learning Algorithms in Estimating the Unemployment Rate

Enes Filiz
Show PDF Cite
Unemployment affects the whole country deeply. It is more than a personal struggle. This rate serves as a key sign of economic health. Governments focus on this issue closely. Low unemployment means the economy works well. It supports a better life for everyone. Many countries offer help to create new jobs. This research predicts future unemployment in Türkiye. It applies RBF Regression, SVM Regression, and Random Forest methods. The study examines data between 2005 and 2025. It tests each algorithm against the others. SVM Regression stands out among them. It reaches a high accuracy of 91%. The study also uses the ReliefF algorithm to pick important factors. Industry and construction sectors matter the most. Employment and labor force participation rates follow them. The results stay strong even with fewer variables. This proves that a simple model works effectively. Successful forecasting requires only the most relevant data.

More from our Archive