Retention Is All You Need

Explainability matters for HR decisions: The paper demonstrates that machine learning predictions about employee attrition become more useful when paired with explainable AI (SHAP), allowing HR departments to understand why a model predicts an employee might leave, rather than just receiving a yes/no prediction.

Individual-level insights enable targeted interventions: By analyzing what-if scenarios for specific employees, the HR-DSS system identifies which particular factors (such as salary, workload, or advancement opportunities) are driving each person's attrition risk, enabling HR to design personalized retention strategies rather than one-size-fits-all approaches.

Transparency supports fairer HR decisions: Using natural language explanations generated from the model's reasoning helps HR professionals understand and potentially challenge the system's recommendations, reducing the risk of opaque algorithmic bias in employment decisions that could unfairly disadvantage certain employees.

Actionable insights over black-box predictions: The paper shows that identifying "dominant features" for each employee allows organizations to take concrete steps—adjusting specific conditions for individuals—to convert predicted attrition into retention, making AI a practical tool rather than merely a diagnostic one.