M.S. Applied Data Science - Capstone Chronicles 2025

2

This project uses exploratory data analysis (EDA) and machine learning techniques to predict employee turnover and identify at-risk employees who were more likely to leave their jobs. Using key drivers (e.g., job satisfaction scores, salary, job level), a predictive model was trained to classify employees based on their retention likelihood accurately. The model enables HR teams to take proactive actions in the work environment to prevent turnover and improve retention rates in the company. 2 Background In HR, employee turnover and retention rates are critical metrics that directly impact the culture and performance of an organization. High turnover can lead to significant costs in recruitment, training, and loss of productivity in the work environment. According to Tatel and Wigert (2024), 51% of U.S. employees were actively seeking a new job. Often, these departures are preventable with timely insights and strategic interventions from HR teams. Vadaman et al. (2015) concluded that a high degree of connectedness decreased the likelihood of an employee’s final decision to leave. Connectedness could vary from organization to organization based on organizational culture and structure. Employee connectedness may be oversimplified and inaccurate to assume employees are only leaving for pay or development. 2.1 Problem Identification and Motivation A primary concern for HR is minimizing employee turnover, which can be costly and disruptive to an organization’s operations. High turnover often leads to increased recruitment and training expenses, reduced productivity, and lower team morale. In many cases, turnover is

preventable with the right insights and interventions. The HR department’s goal is to keep the employee retention rate high so the company can have an overall stable and positive work environment. By taking preventative and prescriptive action to minimize voluntary departures, organizations can improve employee satisfaction, retain high performing and critical talent, and ultimately reduce costs. Preventive action involves early detection of issues by regularly assessing key areas such as: (a) employee engagement, (b) job satisfaction, compensation, and (c) work-life balance. These ongoing assessments help HR teams identify early signs of job dissatisfaction to intervene before issues lead to voluntary exits. Prescriptive actions focus on implementing solutions based on identified risks. Prescriptive actions may include: (a) offering personalized career development plans, (b) establishing targeted retention strategies for at-risk employees, and (c) fostering a culture of recognition, support, and inclusivity. These measures make the organization value its workforce and ensure employee needs are addressed in a strategic and meaningful way. Using preventive and prescriptive actions will ensure the HR team can implement a proactive strategy to maintain high retention rates and reduce the significant costs associated with turnover. 2.2 Definition of Objectives The primary objective of this project was to identify key factors that contribute to employee turnover in an organization. By uncovering patterns and correlations in the data, we aimed to provide actionable insights HR teams can use to implement preventative strategies. These insights

96