Employee Attrition Prediction and Risk Scoring System: An Explainable Machine-Learning Application

This manuscript describes an explainable employee attrition prediction and risk-scoring application developed using Python, Logistic Regression, SHAP explanations, and Streamlit. The application provides a dashboard, employee-level risk profiles, department analysis, and what-if simulations. It is intended as an educational decision-support prototype, not as an automated employment decision system. Formal evaluation metrics and further validation are required before making claims about predictive performance or using the system operationally.