Unemployment Risk of Graduates in Turkey and Europe: Analysis of Educational and Economic Factors Using the KNN Method

This study examines the unemployment risks of graduates with advanced education in Turkey and European Union member states using a comparative approach. The research investigates the effects of economic, educational, and demographic factors — including GDP per capita, income inequality (Gini index), government expenditure on education, labor force participation rate, and school enrollment rate — on graduate unemployment rates for the period 2017–2021. Data obtained from the World Bank were analyzed using the K-Nearest Neighbor (KNN) algorithm. Descriptive statistics, correlation matrix analysis, histogram, Q-Q plot, and boxplot methods were applied during the data preparation phase. Model performance was evaluated using RMSE, MAE, MSE, and R² metrics across K values ranging from 1 to 20. The optimal model performance was achieved at K=2, with an RMSE of 0.2498 and R²=0.929. Findings indicate that Turkey lags behind European countries in key indicators such as labor force participation, education expenditure, and GDP per capita, which contribute significantly to graduate unemployment. This study provides a data-driven perspective for policymakers aiming to reduce graduate unemployment through targeted education and employment strategies. Keywords: Graduate unemployment, K-Nearest Neighbor (KNN) algorithm, labor market, education expenditure, income inequality, Turkey, European Union