Artificial intelligence–based employee turnover forecasting as a decision support tool for HR management

Abstract

In modern organizations, employee turnover is a serious managerial and economic problem that leads to significant financial, operational, and organisational costs. Therefore, understanding the reasons for employee turnover and identifying potential risks at an early stage are critical tasks for human resource management and strategic decision-making. This study proposes an approach to predicting employee turnover based on artificial intelligence using machine learning as decision support tools for HR managers. Instead of focusing solely on the accuracy of predictions, it was decided to focus on the ability to correctly predict employee resignations in conditions of class imbalance, as well as on the balance between model false positives and the model’s ability to identify minority classes. In addition, the study emphasizes the interpretability of models using explainable artificial intelligence methods, which allows managers to better understand the factors affecting employee turnover. The study also used a non-standard approach to provide more reliable model evaluation results, namely Repeated Stratified K-Fold Cross Validation. Finally, the results demonstrate how AI-based predictive models can support HR decision-making by providing transparent and actionable insights into employee retention risks. The findings highlight the potential of explainable artificial intelligence to improve evidence-based human resource management, reduce turnover-related costs, and increase organisational performance. From an economic and business management perspective, the proposed approach contributes to more informed strategic human resource management decisions and supports sustainable organisational performance.