Research: Identifying and Modelling Key Predictors of Employee Attrition Using Machine ...

Research question: Which organisational and individual factors most accurately predict employee attrition, and how can predictive modelling enhance retention strategies?

Employee attrition poses significant challenges for organisations, impacting productivity, morale, and financial performance. With the increasing availability of people-analytics data, companies are seeking advanced tools to proactively identify at-risk employees and mitigate turnover risks.

While prior studies have examined correlates of attrition, there is a research gap in the application of robust predictive modelling techniques to organisational datasets, especially in integrating diverse drivers such as engagement, performance, and DEI metrics. The specific combination and strength of predictors remains underexplored in many real-world contexts.

This study will undertake a systematic literature review, construct hypotheses around potential attrition drivers, and apply statistical and machine learning models (e.g., logistic regression, random forests) to a large-scale HR dataset such as the IBM HR Analytics Employee Attrition & Performance dataset. The expected contribution is a defensible model of retention drivers, offering actionable insights for HR policy and targeted interventions.

Understanding and accurately predicting employee attrition is crucial for organisational sustainability and strategic talent management. This research will support evidence-based HR practices, helping organisations minimise costly turnover and enhance employee engagement.