PREDICTING LONG-TERM UNEMPLOYMENT CASES AND IDENTIFYING CONTRIBUTING FACTORS: A DATA MINING APPROACH

Predicting long-term unemployment cases and identifying factors contributing to them could enable policymakers and employment service offices to take proactive steps in developing effective interventions and support programs tailored to specific groups most affected by long-term unemployment. In this study, we analyze a real-world dataset of unemployment claims in Alabama by utilizing data mining techniques. Specifically, we explore multiple models of logistic regression to predict long-term unemployment cases. The predictive accuracies of the models are compared and discussed. Our analysis also identified the factors contributing to long-term unemployment that include industries with high rates of layoffs, age, monthly wage, and workforce region.