M.S. Applied Data Science - Capstone Chronicles 2025

3

3.2 The Impact of Employee Turnover and Turnover Volatility on Labor Productivity: A Flexible Nonlinear Approach De Winne et al. (2019) study indicates there is a practical significance between turnover and organizational performance. The results showed low levels of turnover benefit organizations regarding labor productivity De Winne et al. 2019). The article stresses the importance of not eliminating turnover, which may create unnecessary cost, but managing and keeping turnover low to balance financial performance (De Winne et al. 2019). 3.3 Overqualified Employees’ Actual Turnover: The Role of Growth Dissatisfaction and the Contextual Effects of Age and Pay Overqualified employees are attractive to employers as they add diverse value and leverage to companies and possess more knowledge, skills, and abilities than the job requires. This makes it crucial to understand why these employees voluntarily leave the organization. Gaps in this analysis are factors and scales limited by the objective assessment of satisfaction (Mah et al., 2025). 3.4 Employee Turnover Prediction Based on Ensemble Learning DGNK Model The researchers experimented with five different machine learning methods to determine an approach that yielded the best accuracy. After evaluating the Area Under the Curve (AUC) values, precision, recall, and F-1 scores, it was concluded the ensemble learning model had a practical application value for the prediction of employee turnover intention. The two-layer ensemble learning model resulted in an accuracy of 95.58% (Ma et al., 2024). 3.5 Predictors of Turnover Intention in U.S. Federal Government Workforce: Machine

support efforts to reduce voluntary exits, improve employee satisfaction, and increase overall retention rates. By identifying these drivers, we can build a model that predicts which employees are most at risk of leaving and why. Some factors may be more important than others. Some may believe employee disengagement leads to turnover and others would emphasize pay or benefits. To achieve our data-driven approach, we developed a machine learning model that can accurately predict whether an employee is likely to stay or leave. Additionally, we segmented employees based on their risk levels, enabling HR teams to implement targeted retention strategies for those most at risk. This data-driven approach will help organizations make informed decisions and develop a more stable and positive work environment. 3 Literature Review The purpose of the literature review was to understand research on the key drivers of employee turnover and predictive models to predict employee turnover. By reviewing recent studies on turnover, the literature review provides background on predictive models used for this task, defines turnover, factors that contribute to turnover, and some gaps in the literature that our project may be able to address. 3.1 Is Turnover Intention Static or Dynamic? The Impacts of Inter-role Conflicts and Psychological Workplace Strain on Turnover Intention Trajectories Turnover intention is an employee’s conscious, deliberate willingness to leave the current organization. Comparatively, turnover is the actual result of an employee quitting. It is important to understand the intentions or why employees decide to finally quit before they do. This research noted gaps in factors between organizations and industries (Jeong & Lee, 2023).

97