Human Resource Analytics: Leveraging Machine Learning for Employee Attrition Prediction

Abstract

Organizations in the 21st century face a major challenge in the form of employee attrition. Due to increasing prevalence of highly skilled workers, amidst the competition among organizations to secure talented employees, finding effective employee satisfaction solutions has become more important than ever. Developing a strong workforce for global competition requires employees in an organization should see themselves as important and indispensable members of the group. In order to predict which factors, contribute to an increase in churn rate, companies must invest in powerful predictive mechanisms. A certain degree of accuracy may enable the organization to identify these factors and eliminate them or, in the worst case, decrease their impact. With machine learning, large datasets can be analysed more efficiently to gain meaningful insight into their complexity. As machine learning platforms have grown, human resource analytics have risen to help managers make some of the most difficult decisions. Organizational ecosystems have improved, resulting in profit growth, stronger organizational ecosystems that have never been seen before a decade ago. Over the top Fortune 500 companies, machine learning algorithms are now standard practice in human resource departments, and similar solutions are rapidly being introduced by other organizations to improve the satisfaction of employee. The best efficacious machine learning algorithms namely Logistic Regression, K Nearest Neighbour, Random Forest, Adaboost and Gradient Boosting for predicting voluntary turnover are analysed in this work to explain the underlying causes of employee attrition.